Abstract: In the process of exploration and development of oil and gas fields, drilling is an important means to obtain geological parameters, research the characteristics of oil reservoirs, and extract oil and natural gas. Due to the heterogeneity of the reservoir, the large difference in the production status of each well, and the complex terrain on the ground, the deployment of well locations has a great impact on controlling the risk of drilling operations, reducing the difficulty of stimulation and reconstruction, and improving the economic benefits of oil and gas wells. Drilling well location optimization is of great significance. Aiming at the difficulty of multi-well well location optimization under constraints such as heterogeneous reservoirs and avoidance areas, and the lack of application tools and software, the Nelder-Mead simplex optimization method was improved by adding constraint items and half-way reflection points. Through continuous iteration, a well location optimization model was established with the cumulative production of the same production time as the optimization goal. The results show that the improved Nelder-Mead simplex well location optimization method can greatly improve the accuracy of well location selection. In heterogeneous reservoirs, multi-well well location selection can control the entire reservoir more uniformly. If there is an avoidance area, the well location adjustment can take into account both the surface and the control of the reservoir to increase the cumulative production.