The application of traditional Kalman filtering is limited when the state of dynamic systems is constrained.A Multiscale constrained filtering method is proposed in this paper to deal with these constrained problems which is described by equation.The state of systems is rewritten as blocked form and wavelet transform are used to obtain new dynamic systems containing constrained information.The wavelet coefficients are estimated with measurements and finally obtain the estimate results of original state by using inverse wavelet transform.New algorithm is capable of Multiscale analysis and the estimate process is real time.Computer simulations show that the trace of the estimate error covariance of the new algorithm is smaller than that of the traditional Kalman filter

and the estimate accuracy of it is better than that of the existing constraint algorithm.