System Design of Computer Simulation Environment based on Deep Learning Algorithm

Nowadays, due to the rapid development and application of deep learning, its application in computer simulation environment system has become a hot topic. The environment model building module in computer simulation environment system can be combined with deep learning algorithm to automatically identify and construct the environment. The real-time optimization module reduces the scale of the model and speeds up the reasoning by pruning and quantization. Under the same simulation conditions, the traditional method and deep learning algorithm are used to carry out simulation experiments, and the output results are recorded. The error rate of the traditional method is 0.15, the mean square error is 0.008, and the accuracy rate is 85%. The error rate of the deep learning algorithm is 0.10, the mean square error is 0.005, and the accuracy rate is 90%. The deep learning algorithm in this paper is helpful to improve the accuracy of computer simulation environment system.