Search

Patients arrive in burn centers with many etiologies of burn incidence (such as scald, contact, flame, steam, etc.). Treatment of these injuries depends on the burn depth and surface area of these burns. Moreover, earlier diagnosis and treatment of these burns lead to better patient outcomes. In this study, we show that THz-TDS imaging with a neural network classification algorithm can reproducibly classify burns, independent of their etiology, with an ROC-AUC of 88% in a porcine model.

The accuracy of clinical assessment of partial-thickness burn injuries has remained as low as 60% in the first few days post burn induction. Here, we present the implementation of a wavelet Shannon entropy technique for noninvasive characterization of burn injuries in an in vivo porcine burn study. Supervised machine learning using the support vector machines (SVM) based on the energy to Shannon entropy ratio (ESER) in the wavelet packet transform of the THz-TDS waveform yielded accuracy rates above 91% in differentiation between superficial, intermediate, and full-thickness burn categories.

Automatic fault detection, localization, and repairing have always been a research hot spot. Because software faults can appear anywhere in a software, it is impossible to automatically repair all types of faults. One possible solution is to detect, localize, and fix a specific fault each time according to the fault signature pattern. Along this direction, we propose an approach that detects pairwise-constraint software faults using algebra graph representation. The approach takes a program and a pairwise-constraint as inputs and generates a fault report for programmers. It firstly generates a constraint control flow graph ( CFG ) that is labeled using pairwise constraint and then translates the constraint CFG as a path expression using algebra graph representation, which is an abstract path model for the program. Finally, it employs a detection algorithm to detect whether the program contains pairwise-constraint faults. We perform case studies to validate the effectiveness of our approach. The preliminary results show that the approach can detect pairwise-constraint software faults before software testing.

Safety-critical systems are those systems whose failure would cause significant property loss or crisis. These systems possibly contain complex predicates, which represent the conditions of system state transformation. Predicate testing is critical to ensuring system safety. However, it is difficult to generate appropriate test cases for complex predicates. To resolve the issue, predicate testing criteria based on major clause are proposed. Major clause of predicate is a clause that directly determines the value of predicate, and the preponderances of testing criteria based on major clause is that not only guarantee the rigor of testing, but also reduce the testing cost. This paper presents an approach that using different test criteria based on major clause to automatically generate corresponding test case constraint set for predicates in safety-critical systems. Firstly, the complex predicate is modeled as an abstract syntax tree (AST). Secondly, truth table for predicates is generated by extracting AST and further filtered based on predicate testing criteria (such as PC, RACC and CACC). Finally, testing case constraints set is generated by a random algorithm to satisfy certain test criteria. Our experimental results show that the constraint set generated by our method is reasonable and our approach is correct. By introducing variants into the empirical system, our approach shows the validity of different test criteria.

A new type of microfluidic platforms, photothermal bar-chart chip (PT-Chip), has been developed using the on-chip nanomaterial-mediated photothermal effect as the novel tunable microfluidic driving force to drive ink bar charts in a visual quantitative readout fashion. This is the first time to exploit the nanomaterial-mediated photothermal effect for microfluidic pumping and also serves as the first report of the photothermal bar-chart microfluidic platform with wide microfluidic applications, especially for visual quantitative point-of-care detection.

In this paper, we propose a smoothing fast iterative shrinkage/thresholding algorithm for compressed MR image reconstruction. The basic idea of our algorithm is to apply a smoothing technique to the minimization problem of a linear combination of three terms corresponding to a least squares data fitting, total variation and $l_{1}$ norm regularization, which has been shown to be very powerful for the MR image reconstruction, and then minimize this approximation using the well-known fast iterative shrinkage/thresholding algorithm. Experimental results show that the quality of restored MR images by our proposed method is competitive with those restored by the previous methods for compressed MR image reconstruction.

Though most of nature terrains, such as flat plains and hills, can be represented well by traditional grid-based DEM, a lot of Nature-artificial Terrains (NATs), like typical terraces in loess plateau in northwest China cannot be represented well and efficiently directly by grid-based DEM. Therefore, the concept, data model and data structures of a new Terrain-objects Oriented DEM called TODEM are presented in this paper. The classification of terrain objects and their encapsulation are discussed in detail. The main attributes and methods of several core classes of TODEM, such as SubTIN, GlobalGrid, Terrace, and TODEM, are also described in details. The methods of constructing TODEM by embedding the SubTINs into GlobalGrid seamlessly are also put forth and used to represent the typical NATs like terraces in loess plateau successfully in our case study. The results of the case study show that the planar and vertical shapes of terraces can be represented well and visualized realistically by the TODEM. The methods of constructing TODEM by using feature lines of NATs are feasible.

Aiming at nonlinearity and large dead-time of object to enhance dynamic quality of closed-loop system, this paper applies single neuron PID controller based on supervised Hebb learning algorithm, combined with correction of actual weighting coefficient, utilizes adaptive and self-learning ability of single neuron, tunes weights of controller and compensate for delay time, in order to improve its dynamic quality of closed-loop system. The simulation result shows that the method given in this paper can achieve good control characteristic and eliminate the impact on dynamic quality of system caused by delay time.

Pulse-taking is one of the very important skills of examining disease in traditional Chinese medical(TCM), but during the pulse-taking process, doctor's subjective factor effects a lot, It's hard to make pulse description to normalization, standardization, digitization, This thesis plan to design a three-dimensional force sensing pulse sensor which is convenient for checking the axial tensile force and radial pulse force at the same time, to perceive information of pulse, also to do some basic work to achieve pulse objectification and standardization of teaching, in this paper, through dynamic test and analysis and pulse measurement experiments to prove its use value, and confirmed with higher accuracy and practicality.

The fault signal of rolling bearing has the feature of nonstationarity. To apply wavelet packet to decompose and reconstruct the vibrational signals collected from testing at first, then the Hilbert operator was applied to achieve the purpose of demodulating the envelope. Finally, the signals after refining envelope spectrum will expand the information of defects. And then extracting the characteristic frequency of fault signals and judging the fault location of the bearing. The experimental analysis shows that wavelet packet and Hilbert operator could be applied to extract the characteristic defect frequency of rolling bearing effectively, and in the other word this method is effective to detect the class of local damage of rolling bearing.

Thermal power plant is a plant which uses thermal energy generated by burning of fossil fuels such as coal, oil and natural gas for generation of electrical energy. But these fuels not only consume energy, but also pollute the environment, so more and more thermal power plant started to use straw as a new fuel which is economic and environment-friendly. In this paper, a microwave moisture measurement method is proposed because of the moisture content is difficult to measure when thermal power plant used straw as a new fuel, and its principle is analyzed and deduced. According to the derivation of the formula, a kind of straw moisture content measurement system is designed. It is convenient to measure moisture content in straw.

In this paper, the prison monitoring system is designed based on wireless sensor network technology. This system could not only monitor the relevant information about the vital signs of inmates in prison, but also record and check the physical condition of personnel, activities, status and location information at any time, preventing inmates from committing suicide or escaping, and discovering whereabouts of the criminals once the jailbreak occurred. The application of wireless sensor network technology in the prison monitoring system enables the monitoring of the prison more comprehensive, more scientific and more effective.

These special characteristics such as distributed, multi-hop, self-organizing, dynamic topology, time-variant channels, and limited resources etc make the traditional routing algorithms or protocols for wired networks and centered wireless networks can’t be used in mobile ad hoc networks directly. So the appropriative multicast routing algorithms or protocols for mobile ad hoc networks must be designed. In the paper, Combining merits between the mesh-based and the tree-based MANET multicast protocol, we proposed ADMRP (Adaptive Distributed Multicast Routing protocol), whose node can computes forward list dynamically and distributed. This mechanics can save the limited bandwidth, decrease the channel collision and reduce the total overhead of the MANET. At last, the ADMRP is validated by OPNET.

A novel method is proposed to design PID controllers for stable first order and high order systems with time delay. Using a first-order Pade approximation for the transport delay, this method is based on developing a certain relationship between the coefficients of corresponding powers of s in the numerator and those in the denominator of the closed loop transfer function for FOPTD processes. Such relationship between these coefficients is searched using particle swarm optimization (PSO) algorithm. Then, the approach is applied to high order systems by approximating them to FOPDT models. The proposed method originates from processes with small time delay; however, it is still effective even if the time delay is quite large. Simulation results show that the proposed method gives significantly better dynamic performances than IMC-PID method and the formula proposed by Smith, C.A. et al.

The miniature unmanned helicopter exhibits a complex and nonlinear dynamic behavior, open- loop unstable and a high degree of inter-axis coupling. This paper describes a hybrid PID velocity control method based on single neuron for an unmanned helicopter. This method possesses both the simple structure of PID controller, and adjusting papameters on-line by intelligent technique. The application results of an unmanned helicopter show that the proposed controller can have helicopter achieve stable attitude in hover and improve the stability of the attitude control system. And the fine robustness for imprecise helicopter model is accounted for.