Search

Interventions from the lairage to the chilling stage of the pig slaughter and dressing process are important to reduce microbial contamination of carcasses and, therefore, constitute an essential part of meat safety assurance systems. A systematic review and meta-analysis of literature investigating the efficacy of interventions during primary processing to control microbiological contamination on pig carcasses was performed. The main objective was to assess the effectiveness of abattoir interventions in reducing aerobic colony count (ACC), Enterobacteriaceae, generic Escherichia coli and Yersinia spp. on pig carcasses. Following a structured systematic review process, a total of 22 studies, which were judged as of low risk of bias, were used for detailed data extraction and meta- analysis. Of these, there were 16 studies conducted in commercial abattoirs, two in research plants and five studies conducted under laboratory conditions. Regarding study design, eight papers reported controlled trials, five reported challenge trials and twelve papers reported use of before-and-after study design. Lairage interventions for live pigs were investigated in only two studies, pre- chill carcass interventions were investigated in 17 studies, carcass chilling in eight and multiple interventions in five studies. The summary effects from a random-effect meta-analysis model were generated. Several commercial trials found that pig carcass scalding was effective in reducing the prevalence of Enterobacteriaceae (RR 0.05) and ACC (2.84 log10 CFU/cm²). Similarly, a significant reduction of these two groups of indicator bacteria on carcasses was found after singeing (RR 0.25, and 1.95 log10 CFU/cm², respectively). Rectum tying was effective in reducing the prevalence of Y. enterocolitica on pig carcasses (RR 0.60). A multiple hurdle approach that included a sequential application of carcass interventions significantly reduced Enterobacteriaceae prevalence (RR 0.11) and ACC on carcasses (2.85 log10 CFU/cm²). Overall, the results suggest that scalding, singeing, washing with hot water and lactic acid, and dry chilling are effective in reducing indicator bacteria on pig carcasses. The meta-analysis also found evidence that pathogenic Y. enterocolitica on pig carcasses is effectively reduced by the standard procedure of tying the rectum ; however, this was the only intervention for Yersinia investigated under commercial conditions.

This paper presents a new approach for improving performances of magnetic levitation system. Controlled parameter is the amplitude which levitated object achieves during movement from one levitation position to another. Two position levitation with improved amplitude performances is obtained by implementing orthogonal neural network in standard levitation control logic. Proposed network is a nonlinear autoregressive neural network with newly developed activation function based on orthogonal polynomials. Performed experiments on a system with default control logic showed that it could not provide stable two position levitation when specified amplitude of the levitation object is greater than 10-4 m. Artificial network was trained using real experimental data and it was based on standard tangent and sigmoid activation functions. Default activation functions were then substituted with a newly developed orthogonal polynomial functions. The amplitude 10-3 m was achieved with stable two position levitation after parameter optimization. It is proven that simple control logic with nonlinear autoregressive neural network and proper activation function can provide improved amplitude performances.

Robustness of the differential pulse code modulation system with the first and second order predictor is considered in this paper. Special focus is on a robust stability of prediction filter with regard to predictor coefficients. Generalization of robustness in classical sense is performed and suitable relations for calculating probability of robustness are derived. The proposed method for robustness estimation is used for the first and second order prediction filters in the case of speech signal.

In this paper the approximation of the optimal compressor function using the first-degree spline functions and quadratic spline functions is done. Coefficients on which we form approximative spline functions are determined by solving equation systems that are formed from treshold conditions. For Gaussian source at the input of the quantizer, using the obtained approximate spline functions a companding quantizer designing is done. On the basis of the comparison with the SQNR of the optimal compandor it can be noticed that the proposed companding quantizer based on approximate spline functions achieved SQNR arbitrary close to that of the optimal compandor.

Planning and forecasting wood resources implies a challenging analysis, which has a direct impact on planning human resources, production timeline, as well as stock management of wooden assortments, which requires a complex data analysis taking into account all inputs that define the yield of wooden material. This paper includes an analysis of monthly time series data from 1991 to 2015 which can be characterized as long time dependence data. In recent years, artificial neural networks have become a popular tool for time dependence data treatment. Therefore, a prediction of monthly requirements of treated wood is performed by developing a new type of neural network in this research. The nonlinear autoregressive model with exogenous inputs (NARX) is used as a foundation of a new network. NARX is a type of recurrent neural network which is a very effective tool for approximation of any nonlinear function, especially ones which could occur during a nonlinear time sequence prediction. The main contribution of this paper is the introduction of an artificial endocrine factor inside the standard NARX structure. The developed ENARX model provides an extra sensitivity of the network to environmental conditions and external disturbances, as well as its improved adaptive capability. The proposed network shows better forecasting performances compared to the default NARX network, thus establishing itself as an excellent prediction tool in the field of wood science, engineering and technology.

Design of a controller that is capable to control the process in order to obtain desired overall system performance is a standard problem in all industries. Use of computers and microprocessors have resulted in increased interest in research and design of the control in discrete-time. The main goal is to obtain optimal control moves in presence of disturbances despite inaccuracies in the model of the system. This implies the need for a robust controller as well as a control algorithm that is not time consuming to compute. The research described in this thesis addresses the combination of the strengths of two di erent control methodologies to design an optimal and robust controller with insignificant online calculation conditions subject to defined objectives. The first approach presents the combination of generalized predictive and sliding mode control (SMC) techniques in order to improve the system robustness to parameter variation. The proposed control algorithm belongs to the group of chattering-free sliding mode control laws, and it provides the minimum value of the cost function in the presence of parameter perturbations. Digital simulation results are given to verify the sliding mode based generalized predictive controller. The next approach presents Tube Model Predictive Control (TMPC) with a SMC as an auxiliary controller. It is shown how to calculate the tube widths under SMC control, and thus how much the constraints of the nominal MPC have to be tightened in order to achieve robust stability and constraint fulfilment. Digital simulations and real-time experiments are performed for demonstrating theoretical framework. Improved TMPC with a SMC as an auxiliary controller is studied and presented in the sequel. To ensure the reduced size of the optimization problem, while allowing for large prediction horizons, orthogonal functions are used for the design of the nominal Model Predictive Control (MPC). The MPC component is parameterized using Laguerre functions in discrete-time. This control algorithm is also veri ed using real-time experiments. Finally, Predictive Sliding Mode Control (PSMC) that uses Laguerre functions in the design of a control input signal is investigated. Two types of PSMC algorithms are considered: one originating from the digital equivalent control method approach, and another containing an additional sliding mode control component that provides the robustness and determines the system dynamics in reaching mode. A one-step-delayed disturbance estimator is introduced to account for system nonlinearities and unknown disturbances, as well as to ensure better system steady-state accuracy. The proposed algorithms are demonstrated by conducting several real-time experiments. Robustness of the closed loop, affected by tuning parameter values, is demonstrated as well. Rezime Projektovanje regulatora koji sluzi za upravljanje procesom u cilju postizanja zeljenih performansi celokupnog sistema je standardni problem u svim granama industrije. Razvoj ra£unara i mikroprocesora je u zna£ajnoj meri uticao na pove¢anje interesovanja za istraziivanje u oblasti projektovanje regulatora u diskretnom domenu. Glavni cilj je dobijanje optimalnog upravljanja u prisustvu poreme¢aja bez obzira na ta£nost modela procesa. To podrazumeva robustnost regulatora kao i to da algoritam upravljanja nekoristi previse procesorskog vremena za izra£unavanje. Istraºivanje koje je opisano u ovoj tezi prikazuje kombinaciju dobrih osobina dveju metodologija u cilju projektovanja optimalnog i robusnog upravljanja sa bezna£ajnim uslovima online izra£unavanja pri unapred definisanim ograni£enjima. Prvi prilaz pokazuje kombinaciju generalizovanog prediktvnog upravljanja i upravljanja sa kliznim rezimima u ciju poboljsanja robustnosti pri promeni parametara sistema. Predlozeni algoritam upravljanja pripada grupi zakona upravljanja sa kliznim rezimima koji smanjuju £etering i obezbe uju minimalnu vrednost kriterijumske funkcije pri parametarskoj perturbaciji. Rezultati digitalne simulacije su dati u cilju provere upravljanja generalizovanog prediktivnog upravljanja zasnovanog na kliznim rezimima. Slede¢i pristup prikazuje Tube model prediktivno upravljanje (TMPU) sa regula torom zasnovanim na kliznim rezimima kao pomo¢nim regulatorom. Pokazano je kako se racuna sirina 'cevi' (eng. Tube) pri upravljanju sa kliznim rezimima, i za koliko treba smanjiti ograni£enja za nominalni model prediktivni regulator (MPR) kako bi se postigla robustnost i stabilnost sistema pri ograni£enjima. Digitalna simulacija i eksperimenti u realnom vremenu su izvrseni kako bi se potvrdili teoretski dobijeni rezultati. Poboljsan TMPU sa regulatorom sa kliznim rezimima je proucavan i prikazan u nastavku. U cilju smanjenja optimizacionog problema, a da se pri tom kosristi dovoljno veliki horizont predikcije, za projektovanje nominalnog MPRa kori²¢ene su ortogonalne funkcije u diskretnom domenu. MPR je projektovan pomo¢u diskretnih Lagerovih funkcija. Ovaj upravlja£koi algoritma je tako e proveren pomo¢u eksperimenata u realnom vremenu. Na kraju je prou£avano prediktivno upravljanje sa kliznim reºimima (PUKR) koje koristi Lagerove funkcije za projektovanje upravlja£kog signala. Razmatrana su dva tipa PUKR algoritma: jedan koji je zasnovan na metodi ekvivalentnog upravljanja i drugi koji sadrzi dodatnu komponentu upravljanja sa kliznim rezimima, koja obezbeduje robustnost sistema i koja odreduje dinamiku sistema u fazi dosezanja. Estimator poreme ¢aja je uveden kako bi se estimirale nelinearnosti sistema i poreme¢aj, kao i zaobezbedenje boljeg odziva sistema u ustaljenom stanju. Predlozeni algoritam je demonstriran pomo¢u nekoliko eksperimenata u realnom vremenu. Takode, pokazana je irobustnost sistema na promenu parametara. digital fulltext is not avialable