Research On Target Control Method For Batch Process Based On Data

As an important producing process, batch process plays a momentous role in the national economy. The batch optimization control method of data-based is an efficient way to achieve the control of complex batch process, which is difficult to establish the process model or contains a huge number of historical data. By the technique of data-mining to static, estimate and use the data efficiently, the method mentioned above achieve the control of batch production process. Therefore, the study of optimization control method based on data for batch process is of vitally theoretical and practical significance.Based on batch process, the basic principle and related algorithm of model-free adaptive control (MFAC) is introduced in the paper firstly, and the idea is applied to the process level, in order to solve the optimization control of MFAC under the condition of no priori production data. Based on which, as to the process optimization problem with nonlinear constraints, it leads into the penalty function strategy, transforming the constrained optimization problem into unconstrained optimization problems, which improves the optimization control method based on model-free adaptive control and achieves the optimization control of batch process. The simulation results verify the effectiveness of the thought. Through the method of MFAC, the optimization control data of different operating model is accumulated gradually under the condition of none initial data, then is used to establish the optimization control case base. To improve the fitness of the case base, the theory of rough set is used to reduce the original case library. While the working condition is changed, by retrieving the case base to match the target operation condition, it helps directing the production efficiently. Aim to a new working condition, to get an optimization control operation by means of MFAC more efficiently, the case strategy matched highest is used as the initial value of the MFAC. The simulation result improves the effectiveness of the method. Firstly, it evaluates the actual production result produced by case reusing, then to the worse condition of the result, uses the case strategy as the initial value of MFAC to re-optimize the production process and proves the feasibility of the method.From the above, the paper improves the optimization control method of both CBR and MFAC, and solves the optimization control problems of batch process based on data under the condition of only one initial reference value.