Research On Bio-Inspired Intelligent Regulation And Control Algorithm And Its Application

In recent years,with the development of economy and the rapid growth of population,the phenomena of expansion of farmlands,lake reclamation,deforestation,and unstainable land and water management have changed the state of the land surface and confluence conditions thereby exacerbating the occurrence and severity of floods.Reasonable regulation of river basin can deduce the risk of flood and plays an important role in ensuring the economy and people’s livelihood.Therefore,it is an urgent task to strengthen river regulation and conduct precise and timely flow control to mitigate flood hazards.To solve the problem of low precision and poor real-time performance of the existing algorithm for multi-tributary watershed control,the heuristic dynamic programming method is introduced into the water conservancy dispatching control system,and the predictive control algorithm is improved based on the biological network regulation mechanism.(1)In order to solve the optimization problem of water conservancy dispatching optimization with the control ideas,this thesis introduces the geographical background of the Murray River basin in Australia and analyzes the data of the basin.The watershed data is processed after understanding the model.Finally,this thesis analyzes the basin in combination with the water conservancy environment,selects the variables of the basin,and explains the rationality of modeling the basin with the neural network.(2)Based on neural network by biological regulation mechanism,this thesis establishes the prediction model of water conservancy basin.In order to improve the traditional watershed system modeling method,neural network modeling is introduced.To solve the problems of neural network optimization method,an optimization algorithm based on biological regulation mechanism is proposed to improve the structure of neural networks.The improved algorithm has the advantage of dynamically adjusting the optimization process in real time.Finally,the model of watershed system is established based on the optimization algorithm.(3)Based on biological network regulation mechanism and heuristic dynamic programming algorithm,this thesis proposes a heuristic dynamic programming algorithm based on biological network regulation mechanism,and simulates the algorithm to prove the control effect in the aspect of water conservancy dispatching.Firstly,the structure and control method of heuristic dynamic programming algorithm is introduced.It is difficult to accurately deal with complex objects with time-varying and nonlinear characteristics.To realize the optimal scheduling and control of water conservancy system,this thesis combines the biological network regulation mechanism with heuristic dynamic programming algorithm,and proposes a heuristic dynamic programming algorithm based on the biological network regulation mechanism.This algorithm consists of the model network control unit,the action network control unit,the critic network control unit and the central coordination control unit.The control effect of the traditional control algorithm is improved and the optimal scheduling control for the watershed system is carried out through combining the biological network regulation mechanism with heuristic dynamic programming algorithm.(4)Based on biological regulation,operational decision-making and heuristic dynamic programming,this thesis proposes an intelligent dynamic programming predictive control algorithm,then simulates and verifies the control effect in multi-tributary watershed.Firstly,this thesis introduces the structure and control theory of the model predictive control algorithm.Because the parameters and structure of the traditional model predictive control algorithm are invariable,there will be model mismatch and inaccurate control in the progress of the control.Although the algorithm itself has a certain ability to overcome model mismatch,it is still difficult to deal with complex objects with time-varying and nonlinear characteristic.To overcome its limitations,this thesis proposes a predictive control algorithm of intelligent dynamic programming based on biological regulation and operational decision-making,which combines the priority factors of operational decision-making,biological network regulation mechanism and heuristic dynamic programming algorithm.In the prediction part,the algorithm uses biological regulation mechanism to perform real-time online training and correction on the prediction model.In the control part,the algorithm adjusts the controller in real time based on the theory of biological regulation and operational decision.Finally,this thesis simulates the predictive control algorithm in a multi-tributary watershed system to verify the control effect.This thesis simulates Bio-BP algorithm,Bio-HDP algorithm and Bio-int-HDP algorithm and compares with traditional algorithms on MATLAB platform.Through experiments of these algorithms,the effectiveness and superiority of the three proposed algorithms in modeling,prediction and control are proved.And these experiments lay a certain foundation for the optimization of scheduling and control of the multi-tributary watershed in the future.