Biomedical Application of a Random Learning and Elite Opposition-Based Weighted Mean of Vectors Algorithm with Pattern Search Mechanism

It is feasible to increase the comfort level of the paralyzed people with the aid of a biomedical application known as functional electrical stimulation system. With the aid of this system, the paralyzed people can perform movements that are normally difficult for them to carry out making functional electrical stimulation a significant solution for disabled individuals. However, to take the advantage of a functional electrical stimulation system, an appropriate control method must be employed. In this work, therefore, a new control approach is presented by employing a proportional-integral-derivative (PID) controller, a modified integral of time multiplied squared error performance index and a novel enhanced metaheuristic tuning algorithm named multi-criteria-based weighted mean of vectors algorithm (MC-INFO). The tuning algorithm is basically an improved version of original weighted mean of vectors algorithm (INFO) using an elite opposition-based and random learning and pattern search mechanisms. In here, elite opposition-based learning and random learning mechanisms are used for further explorative capability whereas pattern search helps to reach better exploitation. Unimodal, multimodal, and low-dimensional benchmark functions demonstrate the good performance of the proposed MC-INFO algorithm against several other metaheuristic algorithms. The proposed algorithm is used to tune the PID controlled functional electrical stimulation system with the aid of the modified objective function. The overall better capacity of the proposed control method for functional electrical stimulation system is demonstrated comparatively with statistical test, transient and frequency responses using original weighted mean of vectors algorithm, reptile search algorithm, moth-flame optimization algorithm and traditional Ziegler–Nichols-based PID controllers. Further confirmation is provided via comparative assessment against previously reported methods, as well.