Based On Redundancy Repair Online Circuit Evolutionary Design Of Finite State Machine

With the rapid development of electronic information technology,intelligent electronic systems tend to be intelligent and miniaturized,and the scale and complexity are constantly increasing.Evolvable Hardware(EHW)can automatically adjust the internal structure according to changes in the external environment,and is committed to self-organization,self-repair and self-adaptation,especially for special environments such as deep sea,aviation,high temperature or extremely low temperature.Evolutionary hardware is the crystallization of the intersection of many disciplines such as biology,electronics and computer science.Circuit evolution design is one of the main core technologies of evolution hardware.The evolutionary algorithm is a known circuit specification tool,and the field programmable logic gate array is used as the circuit implementation carrier.The circuit evolution design can realize the automatic design of the circuit,and can realize self-organization,self-repair and self-adaptation.The finite state machine is a powerful system modeling tool.It can systematically and systematically design the sequential logic circuit evolution using the finite state machine,which greatly reduces the complexity of the evolution design of the sequential logic circuit and improves the stability and system.Main research contents:(1)Using Cartesian genetic programming,FPGA is used as the carrier,and the logic evolution design platform of logic circuit is designed by using processor and VRC virtual reconfigurable circuit technology.Based on the platform,the input projection decomposition technology can be used to decompose the target circuit into a simple circuit module for evolution,which improves the convergence speed.It is used in the online evolution of finite state machine combination logic circuit,which can evaluate the working state of the circuit in real time.(2)A finite state machine circuit evolutionary design algorithm based on state decomposition technique is proposed,which effectively alleviates the difficulty of circuit evolution design of finite state machine.The state machine is divided into multiple sub-state machines according to the state transition frequency,and the complexity of the finite state machine is reduced.The feasibility and effectiveness of the method are verified by online evolution experiments on the hardware platform.(3)A finite state machine circuit evolutionary design algorithm based on redundancy repair technology is proposed,which solves the problem that the growth of the circuit is slow when the adaptability is small,and the growth rate is seriously slowed when the maximum value is approached.Redundant repair technology can effectively reduce the evolutionary algebra of evolutionary circuits,get the correct target circuit faster,and effectively solve the Stalling effect phenomenon.