The Application Of Intelligent Control On Dense Phase Pneumatic Conveying

Pneumatic conveying system, a system of moving solid particle, is widely used in the industry of power, cement, chemical plant, glass, mining and metallurgy. The low speed, high efficiency dense phase pneumatic conveying equipments can solve the difficulties of the breakage of the materials, the blockage and abrasion of the pipelines and reduce the consumption of gas. But, the system always destroy continuity of production because we can't real-time regulate the material conveying. In order to solve this problem, a new control plan is adopted, that we can control the flux of materials by controlling the pressure of convey.The dense phase pneumatic conveying system is a complicated nonlinear system, whose accuracy mathematic model is hard to establish. At present we need controlling model to continue adjusted the material conveying. Modular neural network has the advantage of automatic decomposition of given task and it's modular training of network. It is superior to non-modular methods in it's high efficiency for a parallel network system structure, easy implementation of additional learning and has strong generalization and stability. So it has good characteristics of real time and robustness. This thesis proposes an architecture and algorithm of Parallel Cooperative Modular Neural Network (PCMNN), which makes it possible to carry out the automatic decomposition of a complicated task and the modular training of the network, and which is experimentally proved to be feasible, effective and to be superior to the non-modular method mainly due to its high training speed and improved network performances, or rather high parallelizability, good generality, easy learning of a additional sample and easy implementation of hardware. It is shown from the experimental results that PCMNN presented in this thesis is characterized by a satisfying approach effect, rapid training speed, proper automatic decomposition of a given task, and a dynamic change in network structure in learning, thus being suited to dense phase pneumatic conveying system and being of great real values.In allusion to the feature of complexity and nonlinear, this paper presents a fuzzy-nearul network intelligent dense-phase pneumatic conveying control system. The fuzzy control rules can be adjusted automatically by BP algorithm. Thus, improve dynamic process of fuzzy control system. The Neural Fuzzy Network hasthe advantages of both of them. Not only can it express approximate and qualitative knowledge, like Fuzzy Logic, but also it has the strong ability of learning and expressing non-linearity. More importantly, its structures of networks have clear physical interpretation to users. The simulation results demonstrate that the proposed controller can effectively improve tracking performance and has good real-time performance, control quality and robustness.