A Reinforcement Learning Control Method for Sewage Treatment

Abstract

The invention provides a reinforcement learning control method for sewage treatment. In a control process of a field controller, an input signal of the field controller is acquired, and an output instruction of the field controller and an output instruction of a reinforcement learning model are combined into a final output instruction for output control, wherein a ratio of a numerical value in theoutput instruction of the reinforcement learning model to a total numerical value of the final output instruction is N, and N is gradually increased from 0 to 100%. According to the method, in the mode of improving a control right of the reinforcement learning model through the progressive ratio, a data collection process and a training learning process of the reinforcement learning model can beeffectively placed in an actual scene, and the processes of accumulating original data and establishing a virtual environment are avoided, so that the cost of performing automatic control in the reinforcement learning mode by an enterprise is effectively reduced and a user can conveniently complete the process of reinforcement learning control from traditional control.

Description

technical field [0001] The invention relates to a reinforcement learning control method for sewage treatment. Background technique [0002] At present, the application of reinforcement learning in industrial control is gradually increasing, but it is generally limited to a specific model. The main reason is that the training of reinforcement learning requires environmental support, and in many sewage treatment control, it is necessary to completely simulate the environment of the real scene. The amount of calculation is far greater than the amount of calculation required for reinforcement learning model training itself, resulting in more losses than gains, and in terms of current enterprise technology development, the accumulation of original data is also very problematic. Contents of the invention [0003] In order to solve the above-mentioned technical problems, the present invention provides an enhanced learning control method for sewage treatment. The enhanced learning...

AI Technical Summary

Problems solved by technology

[0002] At present, the application of reinforcement learning in industrial control is gradually increasing, but it is generally limited to a specific model. The main reason is that the training of reinforcement learning requires environmental support, and in many sewage treatment control, it is necessary to completely simulate the environment of the real scene. The amount of calculation is far greater than the amount of calculation required for reinforcement learning model training itself, resulting in more losses than gains, and in terms of current enterprise technology development, the accumulation of original data is also very problematic

Images