| During the operation of mineral grinding process, the optimal control object is not only to ensure the satisfied tracking performance of control loops, but also to control the technique indices that represent the quality, the efficiency, and the consumption of the product processing into their targeted range by adjusting the control loop set-points on-line. However, due to the difficulty of modeling the accurate mathematical models, the difficulty of measuring the technique indices on-line of mineral grinding process, meanwhile the process is strongly influenced by lots of disturbance, so high requirements are asked for optimal-setting method of basic control loops. In practice, manual control is commonly adopted. However, the manual control can’t adjust the setting points according to the conditions of the operation process timely and exactly, which result to low operation efficiency and unsatisfied control effects. Thus in recent years, the optimal control method aiming at improving the entire operation performance is more and more concerned by our enterprises.A series of advanced control software for optimizing set-points based on real-time optimization (RTO), model predictive control (MPC) have been widely used in chemical process and so on. However, these methods need precise mathematical models, so cannot be applied into mineral grinding process for the difficult of modeling of mathematical models. Now, optimal-setting control algorithms based on expert system, case based reasoning, neural networks, reinforcement learning and so on achieved a better results for the industrial which process models are impossible to obtain. However, offerings based on these algorithms grow slowly, and there is not a proper research platform for mineral grinding process. That is already makes a big disadvantage for the further apply for these algorithms.In order to solve above problems, optimal-setting software for mineral grinding process especially its unique algorithm configuration function is mainly designed and developed. The software is supported by the National Basic Research Program (973) of China under Grant 2009CB320604. The software is shown to be useful and efficient from the experiments conducted with a hybrid intelligent optimal-setting control algorithm and a ADP (Adaptive Dynamic Programming) based optimal-setting control algorithm.The main works of this paper are described as follows:1. This thesis introduces current status of optimal-setting control methods and corresponding optimal-setting software. Based on the analysis of mineral grinding process and control problem, with the analysis of the state-of-the-art optimal-setting software, requirements and analysis of the algorithm configuration function is proposed.2. The overall framework of the algorithm configuration function is designed in this thesis. Then the key techniques such as the definition of algorithm description file, the management of algorithm units, and the auto-execution of the control strategy is presentation and completed in detail. More important, the control strategy is modeled by Petri net, which is a very powerful analysis tool for DEDS (Discrete Event Dynamic System). The effectiveness of the auto-execution is also proved in theory.3. According to the designs of the algorithm configuration, WPF and.NET techniques are used to develop the function in Visual Studio 2010.4. Based on this software, a hybrid intelligent optimal-setting control method and a ADP based on optimal-setting control method are conducted in the operation optimal experiment system for mineral grinding process in state key laboratory of synthetical automation for process industrial. The experiments prove that the Petri net based algorithm configuration function of the software is useful and efficient. |
PDF Full Text Request