The large-scale integration of renewable energy generation into new power systems has led to an exponential increase in operational data. Due to the severe temporal fluctuations and large disturbances of renewable energy, a large amount of bad and malicious data can emerge in the new power system. In addition, these operational data are managed in a distributed manner, leading to the poor performance of traditional centralized state estimation methods in calculation accuracy, speed, and other aspects. To solve this problem, distributed state estimation of new power systems based on edge computing is proposed. Firstly, the shortcomings of traditional centralized state estimation methods are pointed out, and a calculation approach for privacy in distributed state estimation is designed, with coordinated variables as the core. Secondly, based on traditional nonlinear state estimation models, an improved linear state estimation method is proposed to improve computational speed. Next, a multi-objective ant colony distributed algorithm based on edge computing is proposed to realize distributed state estimation. Finally, by taking the IEEE57 system integrated into a real-life new power system as an example, the proposed state estimation algorithm is simulated and verified, and the results confirm its accuracy and high computational speed.