In this paper, based on the matrix model to count the number of elements and judge the symmetry of matrix elements, the nearest neighbor space in high-dimensional space is converted into a symmetry matrix. The boundary extraction is discretized and iterated repeatedly until the fitness satisfies the final condition. The genetic algorithm finds the optimal solution, which is transformed into the fitness function, and the system reliability constraints are processed to solve the cost control problem. According to the results, the cost control cost of the construction project is reduced by 10% based on the matrix model. It shows that the proposed model can promptly solve the existing project cost over-budget situation, which makes the enterprise obtain more economic benefits and enhance the market’s competitiveness.