Application Of Fuzzy Neural Network In The Control Of Palstance Of A Combine Cylinder

qIntelligent control technology is the transformation of traditional industries and is the indispensable technology of developing new products especially the intelligent products. It's also the key to raising labor productivity. Combine is one of the main equipment of modern agricultural machinery. The introduction of intelligent control technology to realize the intelligent control of the Combine operation, it is not only the key to improve the production efficiency and threshing quality, but it also has great significance for the entire agricultural production.The forward speed of Combine is an important factor in productivity and work quality of Combine. Domestic and foreign related studies show that it can improve the work quality and reduce the failure rate if to maintain a constant palstance of threshing cylinder through the control of forward speed of Combine.The main purpose of this dissertation is to choose a suitable intelligent control method—Fuzzy Neural Network to the realization of the constant palstance of threshing cylinder to improve the work quality and reduce the failure rate through the analysis of the power consumption model and the common control methods in the threshing cylinder palstance of Combine.The text contains five chapters, as follows:In chapter one, the text introduces the related theoretical background and developments at home and abroad about this topic, and discuss the existing problems about the control of palstance of a Combine cylinder and the way to resolve it.In chapter two, firstly, introduces the process of founding a power consumption mathematical model of axial threshing cylinder and the simulation software of Simulink in Matlab, then design computer simulation model of axial threshing cylinder. Secondly, introduces the common methods in the control of constant palstance of a Combine Cylinder, and adopted a new control method through contrasted the deficient of the common methods.In chapter three, firstly, it states the principles of artificial neural network and the integration approach of fuzzy system and neural network. Secondly, introduces the principles,structures and algorithms of the fuzzy neural network which adopted in this paper and its application in the control of palstance of a Combine cylinder in detail. Thirdly, presents the improvement of the algorithm of BP network.In chapter four, this chapter mainly simulating the fuzzy neural networks control system of palstance of a Combine cylinder, and achieves the expectable result. At the end of this chapter, it presents the deficient and further research ideas and methods of the design.In chapter five, summarizes the whole paper, and show out some questions about the application of fuzzy neural networks in the control of palstance of a combine cylinder to be resolved in the future and next research direction. The main contribution and research significance of the paper is listed as follow:1. On the basis of literature search and related knowledge learning, firstly, it design computer simulation model of axial threshing cylinder, secondly, present adopted a new control method and the further improvements for the algorithm through contrasted the deficient of the common methods in the control of constant palstance of a Combine cylinder, and simulating the control system.2. The fuzzy neural network adopted in this paper, using the rules to training the structure and parameters of artificial neural network, also using the adaptive self-learning ability of neural network to determine the membership functions and membership of fuzzy rules. In contrast to the common BP network—lack of a theoretical basis of structure and hidden nodes, its spatial structure meaning is clearly and ease to understand the weights code and the amount of calculation has nothing to do with their experience. When introduces more complex rules and conditions, it requires only to prescribe the represent meaning of input and output of the neural network, an no need to change the structure and algorithm of entire controller.3. Through computer simulation, the fuzzy neural network which adopted in this page when in application to the control of constant palstance of a combine cylinder, it responded quickly, has real-time nature and little dynamic error, no steady-state error, no overshoot and oscillation, prove that the fuzzy neural network controller are all justified and feasible. Furthermore, it has highly applied values and widely applied prospect.4. Fuzzy neural network is one of the hot topics of intelligent control technology, it plays most important role no only on industry, agricultural control, but also in medicine, architecture and network areas. This research finds a new application area for the fuzzy neural network technology in modern agricultural machinery. Therefore it has important practical significance in the study of fuzzy neural network.