Multi-mode neural control method of nonlinear system in unknown dynamic environment

The invention discloses a multi-mode neural control method for a nonlinear system in an unknown dynamic environment, relates to the technical field of nonlinear control systems, and aims to explore a dynamic mode recognition mechanism and analyze an internal rule of a relationship between recognition speed and system performance by adopting a multi-mode control method. And the rapid learning and repetition capability of the human to the control action is simulated, so that high-performance and low-power-consumption control of the human-like human is realized. According to limited control modes such as system parameter abrupt change and the like, based on a deterministic learning mechanism, self-adaptive controllers are respectively designed to realize closed-loop stability and tracking performance, an identified control system is dynamically stored in a constant radial basis function network, and a high-performance controller set is constructed; a dynamic monitoring system and a dynamic mode recognizer are co