利用智能优化算法自适应调整系统参数是实现随机共振从强噪声背景中检测信号的前提,针对优化算法易陷入局部最优导致参数寻优效果差的问题,提出一种改进粒子群算法(IPSO),在此基础上实现调参随机共振的信号检测;首先,通过分析双稳随机共振系统的参数-输出信噪比理想模型,得到系统参数变化对输出信噪比的影响规律,并指出这种规律在参数寻优中具有的指导意义;然后,依据上述规律提出在PSO中引入单维度速度反馈机制的改进算法,以消除算法在参数寻优过程中搜索方向的随机性和盲目性,提高搜索精度;最后给出了IPSO调参的随机共振信号检测流程;仿真实验结果表明,与采用PSO等算法相比,基于IPSO调参的随机共振获得了更高的输出信噪比;将所提方法应用在工程实践中,取得了良好的信号检测效果.%Using the intelligent optimization algorithm adjusting the parameters is the premise to realize the signal detection from strong noise background by stochastic resonance.But there is a problem that the optimization algorithm is easy to fall into local optimum and lead to a poor effect of parameter optimization,thus an improved particle swarm algorithm (IPSO) is proposed,on which the signal detection by stochastic resonance is implemented.First,the influence of system parameters of output signal-to-noise ratio of the bistable stochastic resonance is obtained by analyzing the ideal model of them,the guiding significance of which on parameter optimization is pointed out.Then,an improved PSO algorithm is proposed by introducing single dimension speed feedback mechanism to eliminate the randomness and blindness of the search direction in the parameter optimization process,thus enhancing the search accuracy.Finally,the signal detection process of parameter tuning stochastic resonance based on IPSO is given.The simulation results show that,compared with PSO and IPSO algorithm,parameter tuning stochastic resonance based on IPSO can obtain a higher output SNR.The proposed method is applied in engineering practice and has achieved good effect of signal detection.
展开▼
