Study On PH Control System Of Algal Liquid Based On Neural Network

Astaxanthin is widely used in health care products,cosmetics and other fields because of its strong antioxidant capacity,while Haematococcus pluvialis is the best material for astaxanthin extraction.With the increasing demand of prawn cyanin in the market,improving the quality and yield of prawn cyanin has become a research hotspot at home and abroad.As one of the environmental factors affecting the culture of Haematococcus pluvialis,pH value of algal solution has also received extensive attention.There are strict requirements for pH value of algal solution in the cultivation of Haematococcus pluvialis.Too high or too low pH value of algal solution will inhibit the growth of Haematococcus pluvialis.Therefore,in order to improve the yield of astaxanthin,it is very important to study the pH control system.Firstly,the paper explains the reason why pH value of algal liquor affects the culture of Haematococcus pluvialis,and according to the neutralization characteristics of pH value and the law of charge conservation in the process of acid-base neutralization,the static model and dynamic model of pH value control system of algal liquor are established respectively;Then the model,structure and learning rules of the neural network are briefly summarized.Taking BP neural network as an example,the algorithm is derived in detail and the calculation steps are summarized.Secondly,the shortcomings of BP neural network in process control are analyzed.In view of the shortcomings,the methods of additional momentum,adaptive learning rate,quasi Newton,conjugate gradient and LM algorithm are proposed to improve BP neural network;The LM algorithm is optimized by the adaptive adjustment formula of momentum factor and fixed parameters,and then the approximation ability of the improved BP neural network algorithm is simulated and analyzed by MATLAB software.Thirdly,MATLAB software is used to simulate the improved LM-BP neural network PID controller and the traditional PID controller.The simulation results show that the improved LM-BP neural network PID controller designed in this paper has shorter regulation time,no system oscillation,small overshoot and stronger anti-interference ability.Finally,a column photobioreactor was designed according to the experimental results of light intensity and pH value of algal solution in the early stage of cultivation of Haematococcus pluvialis in small environment,the PID controller and the improved LM-BP neural network PID controller are applied to the pH control system of algal liquor,and the experiment is carried out.The experimental results show that the improved LM-BP neural network PID controller has shorter regulation time and higher control accuracy.