Adaptive PID controller for cloud smart city system stability control based on chaotic neural network

Abstract

Stability is an essential component in system control especially with those systems exhibiting a non-linear dynamic behaviour. PID controllers are an integral and most commonly used tuning method for control over non-linear systems due to their simplicity and ease of computational complexity. PID controllers are most useful to systems in real time in which the system control parameters vary very slowly or exhibit uncertainty consequently making prediction very complicated and delayed. Continuous and quick adaptation of PID controllers towards varying input determines the efficiency of the non-linear system and this research paper presents a radial basis neural network model in a chaotic configuration for tuning the PID controller in an adaptive manner. However, because mobile cloud computing inherits some characteristics of mobile computing itself, there may be problems such as network congestion, no signal area and crosstalk, which leads to the frequent failure of computing resources and service cooperation, which makes the availability of mobile cloud computing more difficult than traditional cloud calculate the more severe challenges. The scope of this research work is limited to training the neural network with a chaotic map and time series plots are obtained using the experimental results. The efficiency of the proposed chaotic network model is justified by the evaluation of accuracy and robustness of the model even in presence of disturbances.