A control method for multi-objective tasks

Abstract

The invention is a multi-objective task control method, through one or more neural networks, on the basis of the achievement of the main control object, the secondary control object can be achieved to the greatest extent. The sensor detects the external environment to obtain a set of observed values, which are fed back to the main controller or the first neural network to obtain the first control quantity; the first control quantity is used to achieve the main control goal; the observed The value is fed back to the neural network auxiliary system, and the first control quantity is also transmitted to the neural network auxiliary system. The neural network auxiliary system calculates the maximum achievement of the secondary control target on the basis of not affecting the achievement of the main control target. The second control quantity: output through the second control quantity functional system, and the observed value obtained by the sensor detecting the external environment is fed back to the main controller, and the above steps are repeated.

Description

technical field [0001] The invention belongs to the technical field of automatic control, and in particular relates to a multi-objective task control method. Background technique [0002] Multi-objective control refers to the simultaneous control of two or more interdependent objects. In a multi-objective control system, there are multiple control quantities and multiple observation quantities due to multiple target quantities, and the target quantities vary nonlinearly with the control quantities. The existing multi-objective control scheme adjusts the control system parameters based on a large number of experiments and the experience of engineers, so it is relatively difficult to adjust the parameters. Moreover, the equipment itself is aging or the working environment of the equipment is inconsistent with the laboratory environment, resulting in the control results not being optimal, so the adaptability is poor. [0003] as attached figure 1 In the traditional multi-obj...

AI Technical Summary

Problems solved by technology

However, the equivalent method often needs to pay a higher storage cost and is difficult to adjust. It is necessary to update the member functions in the function collection or each tuple in the lookup table one by one.

Images