Search

The thermal conductivity of silicone rubber is very low and does not meet the required thermal conductive applications, so inorganic fillers are added to increase heat transfer which ultimately improves the thermally conductive path. In this study, silicone rubber/hexagonal boron nitride composite was synthesized by the hydrolysis-polycondensation method to enhance the thermal conductivity of the material while retaining acceptable flexibility properties. The introduction of MQ resin reduced crosslink density, thermal stability, tensile strength, and hardness of the silicone resin composite and improved elongation. The addition of vinyl-based MQT and aluminum-based MQT resins improved the properties of the silicone rubber, while the addition of vinyl-based MQT resin reduced the crosslink density, tensile strength and hardness, and improved its elongation and thermal stability properties. While aluminum-based MQT resin did not have a significant effect on crosslink density, tensile strength, or hardness, it also improved elongation and reduced thermal stability. The high filler concentration of hexagonal boron nitride in the composite enhanced thermal conductivity up to 3.253 Wm−1 K−1, while it reduced tensile strength to 1.248 MPa and elongation to 22% but increased hardness up to 75 shore A. The addition of silicone resin improved the thermal conductivity of all MQ, vinyl-based MQT3 and aluminum-based MQT3 resin composites up to 3.661, 3.962 and 4.817 Wm−1 K−1, respectively. For the same three resins, tensile strength was increased up to 1.274, 1.290, and 1.312 MPa, elongation at break was raised to 125%, 188%, and 150%, and hardness was reduced to 69, 71, and 72 shore A, respectively. The addition of silicone resin also showed an effect on density, volatile content, flame resistance, and volume resistivity.

In this study, a multivariate statistical process control was used to analyze the abnormal samples derived from the deviation of optimum processing parameters. The experimental samples derived from the optimum processing parameters were applied as the optimal historical data to determine the control limit, and then the T2 value was obtained from Hotelling's T2 method. If the T2 value exceeds the control limit, the corresponding sample is considered as abnormal. After that, the Runger, Alt and Montgomery method is used to decompose the abnormal T2 value. Then, each quality characteristic value can be obtained and the corresponding decision tree classifier can be implemented. To improve the classification accuracy, we classify the decision tree classifier into single–double identification, single-factor abnormality and double-factor abnormality. For the individual classification test, the result showed that the accuracy of single–double identification was 98.6%, the single-factor abnormality classification was 100% and the double-factor abnormality classification was 96.0%. For the combination classification test, we can get a 98.6% accuracy rate for the single–double identification, 98.3% accuracy rate for the single-factor abnormality classification and 95.3% accuracy rate for the double-factor abnormality classification. Therefore, it can be confirmed that the proposed methods in this study can effectively identify abnormal samples and establish a fault processing parameter diagnosis system for melt spinning machines.

Melt spinning is the most extensively used method of fabricating polymeric fibers in the textile industry. This series of studies aimed to construct an automatic abnormality diagnosis system for polypropylene (PP) as-spun fiber produced by the melt spinning process. Part I of this study aimed to construct the processing parameter optimization for the PP as-spun fiber produced by the melt spinning machine. The product quality resulting from the processing parameters of the melt spinning process included six control factors: extruder temperature, gear pump temperature, die-head temperature, rotational speed of extruder, rotational speed of gear pump, and take-up speed. The quality characteristics included fiber fineness, breaking strength, breaking elongation, and modulus of resilience. The quality data were derived from the experiments, the design of which were based on the orthogonal array of the Taguchi method in order to calculate the signal-to-noise ratio, analysis of variance, and confidence interval. Principal component analysis was then applied to eliminate the multi-correlation of the output responses and transform the correlated responses into principal components, to obtain multi-quality optimum processing parameters. These optimum parameters, including the extruder temperature (180°C), gear pump temperature (220°C), die-head temperature (240°C), the rotational speed of the extruder (7.5 rpm), the rotational speed of the gear pump (15 rpm), and take-up speed (700 rpm) would later be used to build a prediction of an abnormality diagnosis system for identification of fault processing parameters in a melt spinning machine in Part II of this study.

This study took nano-silica particles mixed in low-density polyethylene to implement the optimum processing parameters and make abnormal parameter identification for the twin-screw co-rotating extruder used in the manufacturing process. The mixing quality was divided into distribution and dispersion, where distribution was tested by an energy dispersive spectrometer and evaluated using the coefficient of variation. Dispersion was assessed by the surface effect and specific surface equations, as based on the spectrum of scanning electron microscopy. By using the Taguchi method in planning the experiment coupled with an analysis of variance, we conducted the single-quality characteristic analysis of the experimental results of the two quality characteristics, namely the distribution and dispersion. Then, by using the hierarchical architecture of analytic level process, we can obtain the optimized parameter factors and levels and the calculation of the total weighting of various parameter levels, as well as the ranking of the parameter levels. According to the confirmation experimental results, the signal-to-noise ratios of distribution and dispersion fell within 95% confidence intervals, indicating that the experiment can be represented and reliable. The optimum parameters combination is SiO2 addition level 1%, screw speed 60 rpm, mixing time 5 min, temperature (upper) 150℃, temperature (middle), 175℃ and temperature (lower) 190℃. After that, by using the optimal parameters and operation processing parameters for support vector machine classification, the abnormality of the processing parameters can be identified for 100%. The good quality of the production can be guaranteed during the extrusion.

The aim of this study was to present a two-step optimization system to find the optimal process parameters of the multi-quality characteristics for transparent polypropylene (PP) composites with hi...

Polypropene (PP) has weak mechanical properties; relevant fibrous composite products are likely to bend under stress, becoming unavailable due to creases and folds. In order to maintain the product haze and tensile characteristic, the toughening property is enhanced greatly in this study. Firstly, the PP and enhanced rubber segment-styrene ethylene/butylene styrene (ERS-SEBS) are analyzed, and the two materials’ melting points and cracking points are confirmed. Three proportions of composite are made by single-screw mixing. The cost is reduced and the efficiency is confirmed by the Taguchi experiment. The single quality optimum combination is obtained by analysis of variance (ANOVA) and a factor response table. The optimum process parameters are designed according to the contribution degrees of quality weight and control parameters by using the analytic hierarchy process and ANOVA of the Taguchi method based on the reproducibility of the single quality optimum combination validation experiment. According to practical validation, in the ERS-SEBS modified optimum process of PP, the impact strength is 7.26 kJ/m2, higher than that of regular PP by 142%. The tensile strength is 23.69 MPa, high than that of regular PP by 3%. The haze can be reduced to 5.7%. The developed composite of the PP/styrene triblock copolymer has better mechanical properties and retains its optical performance. It can be used in a fibrous composite to make a composite with transparent fiber to present the fiber line distribution of fabric in the composite.

Polarizing films are critical components for a wide range of products and their inspection is helpful to enhance product quality. Inspection and classification of the normal and five types of defects for polarizing films are presented using image processing and neural network approaches. The defects are cloud chromatism, strip chromatism, spot chromatism, scratch and poor pasting. Three features, the area, average intensity and compactness, are selected according to the shapes and brightness of the defects regions. The number of training samples are 20, 30 and 40, and the number of testing samples is 40. The results show the recognition rate is 100% when the number of training samples is greater than or equal to 30, proving that the back-propagation neutral network can achieve a high recognition rate with enough training samples, and it can be successfully applied to the inspection of polarizing film defects.

Diagnosing and treating lung cancer at an early stage can improve the survival rate of patients. This study attempted to develop a computer-aided detection (CAD) system. In order to include all nodule types in the detection, this study proposes an image processing method for detecting ground glass opacity (GGO), part solid, and solid nodules in chest computed tomography. The process comprises image preprocessing, lung segmentation, nodule enhancement, candidate detection, and reduction of false positives. For lung segmentation, the edge searching method replaces the computing-intensive iterative hole-filling method. In order to extract nodules with extensively distributed gray levels, image accumulation is used in the nodule enhancement to rapidly enhance the gray level of individual nodules. In order to reduce false positives, the support vector machine (SVM) is applied twice. On the first run, the candidate nodules are obtained by using 4 two-dimensional features, and the classification result is obtained by using 11 three-dimensional features on the second run. This study used 667 lung nodules for experiment and evaluation. The proposed system can detect GGO, part solid, and solid nodules and takes only 0.1 s to process a single image. The total sensitivity of the system is more than 92.05%. The system excels at detecting small nodules in the range of 5 mm–9 mm with a sensitivity of 93.73% and GGO with a sensitivity of 93.02%. The results showed that the proposed rapid detection system has high sensitivity and low false positives, contributing to helping the clinicians’ diagnosis.

Image inspection by wavelet packets and a neural network classifier is presented for non-defect and six kinds of defects in knitted fabrics. The types of defect include a hole, set mark (coarse), dropped stitch, oil stain, streak, and tight end. In this study, wavelet packet decomposition of a sample image is carried out based on the best-basis wavelet packet tree with three resolution levels. The lowest-two entropy among all sub-band images and the standard deviation for the original image are selected as feature inputs of the neural network classifier. These textural features are shown in seven groups, which are separately distributed in the feature space. We gathered a total of 112 experimental samples, with 16 samples in each of the seven aforementioned categories. The results demonstrate that with the three features, 56 test samples are correctly inspected. However, the lack of one of the three features yields wrong classification of some samples. Therefore, the three features selected are definitely suitable for recognition of our knitted fabric defects and also are the smallest number of features required to give accurate inspection.

The current analysis of fabric weave diagrams requires using fabric analyzing glass to record the weave number manually. This method damages eyesight and is also very time consuming. In addition, the unweaving mode damages the weave structure of woven fabric. This study uses a computer vision system and digital image processing technology for direct non-destructive analysis of the commonly used 12 fabric textures of woven fabrics without unweaving. Moreover, it proposes an automated woven fabric weave recognition method to enhance the practicability and fault tolerance of the recognition system. Firstly, the woven fabric image was shot by using a front light source and back light source, the noise of the woven fabric image was reduced by using a median filter and the contrast was increased by using histogram equalization. The statistical threshold value was used to segment the warp yarn area and the opening operation of morphology was used to disconnect the connected blocks and erode small noise. Horizontal projection and vertical projection were used to segment the warp yarn and weft yarn. The weave diagram was drawn to improve the computing time of the gray-level co-occurrence matrix. The contrast in the gray-level co-occurrence matrix was selected as the eigenvalue. In terms of woven fabric samples, 12 target samples were obtained, the Euclidean distance classifier was used and the 12 test samples were used for the experiment. The result showed a recognition rate of 100%. The recognition system was adopted by this study to effectively recognize the woven fabric weave.

This study focused on the abnormal condition of diameter variation in melt spinning, and attempted to recognize the machine processing parameter that deviates from the set value. The experiment used polypropylene as the experimental material. The machine processing parameters included nine factors, including temperatures in three extruder barrel sections, temperature of metering pump, die temperature, spinneret temperature, speed of an extruder screw, speed of metering pump and speed of take-up roll. The biaxial laser sensor measured the yarn diameter instantly, and the optimum parameter combination for minimum diameter variation was obtained by Taguchi method and analysis of variance (ANOVA). The degree of influence of various processing parameters on the diameter variation was determined. The optimum parameter combination was used as parameter setting value, and the processing parameters of various factors were changed for experiment, so as to obtain the signals of abnormal condition. Two methods were used for feature extraction. The minimum entropy of wavelet transform signal was used as eigenvector, and two features were selected by analyzing the statistical process control chart, which are skewness and kurtosis. The abnormal processing parameters and normal condition were recognized effectively and accurately by using the three eigenvalues and back-propagation neural network. With enough training samples, the recognition success rate was as high as 100 %. For complete abnormity diagnosis of melt spinning machine, a two-factor classification process was established by using the single-factor classification result and back-propagation neural network. The experimental results proved that the proposed method can recognize various abnormal conditions successfully and effectively.

In this study, with the glass fiber reinforced Polyoxymethylene composite material as the subject, we examined the impact of different injection molding process parameters such as melt temperature, mold temperature, packing pressure, injection speed and packing time on mechanical properties. We designed the experiment by using the orthogonal array of Taguchi method, and obtained the single quality characteristic optimization parameters by using the main effect analysis and variance analysis of Taguchi method. Based on the experimental quality data, we integrated principal component analysis and grey relation analysis to identify the combination of multiple quality characteristics optimal process parameters. As the research results suggest, if the four quality characteristics including tensile strength, hardness, impact strength and bending strength are considered, the optimal conditions are glass fiber content of 20 wt%, melt temperature 230 °C, mold temperature 60 °C, packing pressure 50 MPa, injection speed 60 mm/s and packing time 1.5 s. Finally, it was verified that the planned experiment of this study can effectively enhance the material’s multiple quality characteristics with good reproducibility.

In the melt-spinning process, evenness of melt-spun yarns will affect the appearance, hairiness, strength, and the production of yarns. Great variations in yarns may cause defects. Yarns with small variation can have a more stable quality. In this study, we applied the fuzzy neural network (FNN) control theory to the melt-spinning machine. By adjusting the speed of the take-up rollers, the average yarn diameter reached the target value, allowing a reduction in variations. Diameter error and diameter error variation were also used as input, and the increments in take-up roller speed were used as output. The FNN was used to adjust the mean and standard deviation of the membership function in the second layer and the connection weighted value of the third and fourth layers in order to achieve convergence and the learning effect. The experimental results showed that the FNN controller maintained the average diameter of yarns and reduced the variation of the yarn diameter. Therefore, the proposed method could be successfully applied to on-line control of yarn evenness.

The laryngeal video stroboscope is an important instrument to test glottal diseases and read vocal fold images and voice quality for physician clinical diagnosis. This study is aimed to develop a medical system with functionality of automatic intelligent recognition of dynamic images. The static images of glottis opening to the largest extent and closing to the smallest extent were screened automatically using color space transformation and image preprocessing. The glottal area was also quantized. As the tongue base movements affected the position of laryngoscope and saliva would result in unclear images, this study used the gray scale adaptive entropy value to set the threshold in order to establish an elimination system. The proposed system can improve the effect of automatically captured images of glottis and achieve an accuracy rate of 96%. In addition, the glottal area and area segmentation threshold were calculated effectively. The glottis area segmentation was corrected, and the glottal area waveform pattern was drawn automatically to assist in vocal fold diagnosis. When developing the intelligent recognition system for vocal fold disorders, this study analyzed the characteristic values of four vocal fold patterns, namely, normal vocal fold, vocal fold paralysis, vocal fold polyp, and vocal fold cyst. It also used the support vector machine classifier to identify vocal fold disorders and achieved an identification accuracy rate of 98.75%. The results can serve as a very valuable reference for diagnosis.

This study melted and mixed poly lactic acid and nano mica into a new nano composite material for injection molding. The injection parameters for the mixture included the melting time, screw speed, heating plate temperature, nano mica content, melting temperature, nozzle temperature, holding pressure, and injection speed. The quality considerations of the new nano composite material were ultraviolet (UV) absorption rate, far-infrared absorption rate, Rockwell hardness, and flame resistance. This study applied main effect analysis and variance analysis theory of the Taguchi method to analyze the single quality characteristic of the experimental results. The optimal factors and level combination was identified through multi-quality integration using the Analytic Hierarchy Process. After consistent examination and calculation of the total weighting of various parameter levels, as well as the ranking of the parameter levels, the optimal injection parameters were obtained. According to the confirmation experiment results, the Signal-to-Noise ratio of the four qualities for the new nano composite fell within the 95% confidence interval, indicating the experiment was reproducible and reliable. The results indicated that the best qualities for the new nano composite are UV absorption rate at 98.22%, far-infrared absorption rate at 90.53%, Rockwell hardness at 108.07, and the limiting oxygen index for fire resistance rating at 30.

Proper setting of process conditions in the melt spinning setup is one way to yield uniform quality of spinline tension. The optimum setting to give uniform spinline tension is determined using experiment plans in the Taguchi method and significant process parameters are also identified. When the setting shifts from the optimum, the spinline tension becomes non-uniform and downgrades product quality. This study aims to diagnose single or double fault conditions of those significant process parameters based on the spinline tension signal. The critical procedures of fault diagnosis are feature extraction and classification. The tension signal is decomposed into a wavelet packet tree of four resolution levels. Four entropies from the best-basis wavelet packet tree and the lowest entropy at level four are selected as features. The back-propagation neural network acts as a classifier. The experimental results demonstrate that the features and classifier actually work well to identify the single and double fault conditions with high accuracy in melt spinning.

In the field of yarn spinning engineering, the importance of the processing parameters taken depends directly on the quality characteristics of the yarn. This study aimed to find the optimal processing parameters for an open-end rotor spinning frame at work to identify its multiple quality characteristics for yarn. In this study, Bamboo charcoal and cotton 70 %/polyester 30 % (CVC) blended fibers were adopted as the materials, and the open-end rotor spinning frame was used to spin the yarn. In order to identify optimal conditions of an open-end rotor spinning frame, the Taguchi experimental method was applied to design open-end rotor spinning experiments, and the L9 orthogonal array was chosen in accordance with nine sets of experiments and contained four control factors and three levels. Furthermore, a response surface methodology (RSM) was used to obtain the models of significant processing parameters for the strength, unevenness, I.P.I, and hairiness. Based on experiments designed to obtain an open-end rotor spun yarn Ne 30, the strength, unevenness, imperfection indicator/km (I.P.I) and hairiness were then chosen as the quality characteristics. In addition, grey relational analysis integrated the optimal processing parameter of multiple quality characteristics, and a confirmation experiment was performed. In conclusion, the optimal processing parameters under steady spinning conditions were a rotor speed of 88000 rpm, a feed speed of 0.392 m/min, and a winding speed of 39.466 m/min.

Image inspection of nine kinds of nonwoven defects by the wavelet transform and neural network is presented. The defects include black yarn, hole, needle streak, oil stain, stripe, corrugation, white spot, folding mark, and wrinkle mark. The wavelet transform decomposes an original image into four subimages in different frequency bands. Four texture measures, the energy, contrast and correlation with gray-level co-occurrence matrices as well as the energy with wavelet coefficients, are selected as defect features and computed based on the low-frequency subimage at resolution level one. The feature values are distributed in groups by the categories of defects throughout the feature space, accounting for suitability of the four features for inspecting nonwoven defects. The subimage is a downsized approximation of the original image; thus, in this manner, feature extraction can not only consume less computation time but also maintain the classification accuracy. The neural network acts as a classifier, which is trained by forty-five samples. The experimental results demonstrate that among forty-five testing samples, the classification accuracy is 100 %.

The uniform tension at a preferred value of polyester (PET) films in processing or rewinding is important to product quality, which can be achieved by tension control. Furthermore, the uniform transport speed is helpful in creating uniform film tension. In this article, a simplified version of PET film processing systems, consisting of an unwinding roll, a sensor roller, and a rewinding roll, is considered. The unwinding roll and rewinding roll are driven by a torque-controlled motor and speed-controlled motor, respectively. A control signal to the torque-controlled motor, generated by a conventional sliding-mode controller or a fuzzy sliding-mode controller, regulates the film tension. Moreover, the speed-controlled motor adjusts the transport speed at the winding section. To improve the speed fluctuation, an external proportional-plus-integral-plus-derivative (PID) controller for the speed-controlled motor was implemented. The experimental results demonstrate that fuzzy sliding-mode control gives the te...

The coefficient of variation (CV) in yarns is the most important evenness characteristic in textile processing and quality control. This article develops a new infrared device for monitoring the coefficient of variation in yarns, based on on-line measurement of the yarn diameter, during the yarn manufacturing process. The device is composed of a 555 astable oscillator, four pairs of infrared emitters and receivers, a summing amplifier, an inverting amplifier, an alternating current (AC) to direct current (DC) converter, a unity-gain second-order Sallen–Key low-pass filter, and a data acquisition system. The optimum values of some factors with the circuit, including the oscillator frequency, amplified gain, cutoff frequency of the low-pass filter, sampling time, and number of sampling data to be averaged in the moving average method, are systematically chosen by the Taguchi method to reduce the variance in the output voltage of the device. The CV based on measured yarn diameter data is transformed to that based on mass profiles. The experimental results reveal that the CVs evaluated by the infrared device are close to those by the Uster Evenness Tester, which is verified by the statistical analysis of variance. © 2007 Wiley Periodicals, Inc. J Appl Polym Sci 2007

Edge alignment of polyester (PET) films is important for achieving product quality and processing speed in winding, coating, drying, and other processes. The edge alignment can be achieved by lateral deflection control, provided that the film tension and transport speed are even at desired values. This article aims to correct the lateral deflection of films by designing robust controllers to swivel the guiding rollers and to maintain even tension and speed at target levels. The self-tuning neuro-proportional integral derivative controller and adaptive high-gain output feedback controller are adopted to guide the lateral deflection so that the film aligns at the desired position. A control scheme, neuron controller by associative learning, is used for maintaining tension and speed control. These strategies are applied to a simplified PET film processing system. The experimental results demonstrate that in our setup, the control schemes can effectively alleviate not only the lateral deflection but also the tension and speed fluctuation at target levels. © 2008 Wiley Periodicals, Inc. Adv Polym Techn 26:153–162, 2007; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/adv.20096

The spinline tension plays a critical role in the development of fiber structures and the quality of as-spun yarns in melt spinning. Implementing a controller to adjust the spinline velocity is helpful to maintain the spinline tension at a target level with small fluctuation, enabling as-spun yarns to possess the desired tenacity and uniform qualities. The spinline tension system is difficult to model and the stochastic disturbance always exists. The discrete adaptive sliding-mode controller can robustly and adaptively deal with the system with the unknown model and stochastic disturbance, such as the spinline tension system. The algorithm estimates the parameters of the controller in the sense of minimizing the deviation from the sliding surface, thus reducing the variation of the tension response about the desired level. The sliding surface is defined by an asymptotically stable polynomial, and seven stable polynomials are chosen in experiments. The experiments are carried out by using a laboratory type of the melt spinning setup to produce polypropylene as-spun yarns. Compared with the results without control, the proposed controller can not only maintain the mean of the tension response close to the target level but also reduce the standard deviation to the value, which is generally acceptable to the manufacturer. © 2006 Wiley Periodicals, Inc. J Appl Polym Sci 100: 3816–3821, 2006

This investigation considers a quantitative procedure for determining the values of critical process parameters in melt spinning to optimize the qualities of denier, tenacity, breaking elongation, and denier variance in as-spun polypropylene yarn. An orthogonal array in the Taguchi method defines the minimum set of parameter-level combinations that are experimentally tested. The significant process parameters, namely the third extruder barrel temperature, spinning temperature, metering pump speed, and take-up velocity, are identified on the basis of the analysis of variance and F test. After a confirmation experiment is conducted to ensure the reproducibility of the experimental results, the back-propagation neural network establishes a continuous system linking 10 process parameters and four qualities. The technique for order preference by similarity to an ideal solution can be used to obtain a performance measure for assessing multiple qualities. The genetic algorithm attempts to find parameter values for optimizing the quality performance, including the denier, tenacity, breaking elongation, and denier variance. Finally, the experimental results demonstrate that the smallest denier, largest tenacity, smallest breaking elongation, and second smallest denier variance of as-spun polypropylene yarn can be achieved with the proposed approach in melt spinning. © 2006 Wiley Periodicals, Inc. J Appl Polym Sci 100: 2532–2541, 2006

An approach for determining parameter values in melt spinning processes to yield optimal qualities of denier and tenacity in as-spun fibers is presented. The approach requires a fewer number of experiments than conventional methods. An orthogonal array in the Taguchi method determines the minimum number of experiment trials to be conducted. Whether the experimental data are adopted to train a neural network is justified by an analysis of variance(ANOVA) and confirmed by experiments. A neural network relating 11 process parameters and two quality characteristics is constructed. The genetic algorithm is aimed at finding parameter values in a continuous solution space to optimize a performance measure on denier and tenacity qualities, based on the neural network. The performance measure is evaluated by the technique for order preference by similarity to ideal solution (TOPSIS). To expand the solution space, three different sets of level values for the orthogonal array are chosen from the ranges where the melt spinning will properly work. The results demonstrate that the proposed approach gives the smaller denier and the larger tenacity of polypropylene(PP) as-spun fibers than the Taguchi method.

This paper presents an experimental study of self-tuning fuzzy control of the two-end loads of padders. A simplified padder set-up is used. In conventional fuzzy control, membership functions and control rules are designed based on human subjective criteria, so very often, the design attempt may not lead to excellent performance. Our membership functions and control rules for load control of padders are experimentally investigated in this application. However, the ability to maintain two-end loads at the desired values is not satisfactory. To improve performance, membership functions are tuned based on an optimization technique of genetic algorithms to yield self-tuning fuzzy control. The experimental results indicate that the self-tuning fuzzy controller can maintain the two-end loads not only evenly and at the desired values. The tuning scheme is helpful in eliminating time-consuming trial-and-error procedures when refining membership func tions or control rules.

One of the factors critical to dyeing levelness in the textile industry is pH control, but it is a difficult problem because of inherent nonlinearity and changing process dynamics. The proportional integral derivative (PID) controller has been widely used in industry. Conventional fixed gains are hard to achieve with dyebath pH control, since it is difficult to model the process and the scheme is not adaptive to time-varying characteristics. Thus, a fuzzy gain-scheduled PID controller is promising for dyebath pH control because of its self-adaptation and robustness. The control scheme uses fuzzy rules and reasoning to autotune PID gains without the need of modeling. A modification to overcome the difficulty in finding the ranges of gains for stability is proposed. Simplified dyeing experiments demonstrate that the proposed controller can track pH set values under the influence of temperature rise, residual alkali, etc.

This paper presents an experimental study of fuzzy self-organizing and neural network control in developing an autoleveling system with a drawing frame. Without the need of modeling, both control strategies can cope with nonlinear or very complex processes, even when subject to random disturbances such as drafting processes. In fuzzy self-organizing control, control rules to improve sliver irregularities are constructed in the basic fuzzy control level. The self-organizing scheme is able to improve the rules automatically. A three-layer neural network model, which approximates the process, is used to compute the Jacobian matrix, which is needed in training the weights and thresholds on-line with the neural network controller. In a laboratory scale of the drawing frame with two drafting zones and two-sliver doubling, the draft ratio is adjustable by regulating the speed of the middle roller. Levelness performance is evaluated by the CV% of sliver products. The experimental results show that both controllers are effective in reducing the CV%, and the neural network controller yields more level slivers than the fuzzy self-organizing controller.

This paper presents procedures to design a minimum variance control for leveling slivers in a double-zone drafting process. We model a random sliver and consider the drafting model as a stochastic system with a zero-mean white noise process and a constant sequence representing the input-sliver mean thickness. Such a model enables the proposed control algorithm not only to smooth out the irregularity of the output sliver, but also to regulate the mean thickness at the desired value. We develop the minimum variance control applicable to our stochastic model based on an optimal predictor. Since some parameters and the noise process are not exactly known, identification is used to refine the model, and the control law is thus designed based on the identified model. The simulation results demonstrate that the algorithm can achieve goals by adjusting the speed of either the back or front roller.

We present an efficient and nondestructive way to compute fiber angles and twist angles of yarns using image processing. The pixels over the floats in an image that can portray fiber paths are detected and labeled, then the labeled pixels are connected to a chain-coded curve by tracing the path of individual fibers. Thus, fiber angles are estimated by the trigonometric relationship between the measured rectangular components of the chain-coded curves. The approximate twist angle of a yarn is statistically computed from all the estimated fiber angles with this yarn. Experiments involve plain, twill, and satin fabrics, each with twenty specimens, and the results show good agreement with manual measurements.

Image classification by a neural-fuzzy system is presented for normal fabrics and eight kinds of fabric defects. This system combines the fuzzification technique with fuzzy logic and a back-propagation learning algorithm with neural networks. Four input features—the ratio of projection lengths in the horizontal and vertical directions, the gray-level mean and standard deviation of the image, and the large number emphasis (LNE) based on the neighboring gray level dependence matrix for the defect area—are selected and their usefulness is justified. The neural network is also implemented and compared with the neural-fuzzy system. The results demonstrate that the neural-fuzzy system is superior to the neural network in classification ability.

A procedure is proposed for decoupling fabric in alignment during processing from all possible deflection disturbances. The dynamics of lateral deflection for a two-span system is mathematically described and refined to compensate for modeling errors through a parameter estimation algorithm. A disturbance decoupling problem with stability (DOPS) is adopted to eliminate the effect of deflection disturbances on a fabric that is to be fed into a process machine in a desired alignment, while the resulting closed-loop system is stable. The simulation results demonstrate that DDPS enables the fabric to stay in alignment by both shifting and swiveling a guiding roller.

pH control has received considerable attention in dyeing processes because of its critical role in quality assurance. In dyeing, pH exhibits strong nonlinearity and time- varying behavior, thus increasing the difficulties of control by conventional means. This necessitates the creation of a controller that adapts itself to changing system parameters, and a new approach involving a self-tuning proportional-integral-derivative (PID) control ler is proposed. The gains of the PID controller are automatically tuned by an estimated parameterized model, which is updated by input and output sequences. The simulation results illustrate that the controller can bring the pH from any initial value to any other value for a simplified dyeing process. The approach is expected to achieve the same results in practical dyeing due to its features of self-tuning and robustness. Hence an operator can manipulate pH control even without knowing the design philosophy.

Dyeing processes are receiving growing attention in the textile industry because of the high demand for quality products. Commonly, dye concentration, pH, and temperature in the dyebath are the factors that most affect shade and coloration uniformity of dyed goods. To achieve quality assurance, control of dye concentration, pH, and temperature is definitely required. Since such factors behave nonlinearly and interact, it is almost impossible to create an exact mathematical model. As a consequence, control by conven tional means is very difficult. One approach to dealing with such a complex process is fuzzy control, which simulates the decision-making activities of experienced experts and employs a reasoning scheme to infer control actions. A small laboratory scale experiment has yielded results showing that fuzzy control is able to control dye concentration, pH, and temperature at the desired values.

A procedure is presented for tension and line speed control in fabric finishing. The dynamics of the tension and line speed is mathematically modeled, exhibiting nonlinear behavior. Based on the nonlinear model, to compensate for modeling uncertainties and unknown disturbances, sliding mode control is adopted because of its robustness and ability to handle nonlinear systems. The proposed approach is implemented for computer simulation and experiment. The results show consistency in performance, and demonstrate that sliding mode control can efficiently accomplish tension and line speed control.