The integration of blockchain and edge computing offers a promising solution to key challenges in Internet of Things (IoT) applications, including data security, transmission reliability, and computility demands. This paper focuses on the total system revenue in IoT application scenarios and proposes a resource allocation method for blockchain networks based on edge computing, which mitigates resource waste by adjusting block rewards. First, an IoT-blockchain architecture supported by the edge server’s computilities is proposed. The revenue models for IoT devices and the edge server incorporating the Gompertz function are constructed, and the resource allocation problem is formulated as a nonlinear programming (NLP) with the objective of maximizing the total system revenue. The Gompertz function effectively characterizes the relationship between block rewards and total computilities, thereby suppressing resource waste caused by IoT devices’ excessive pursuit of block rewards. Then, to address the challenge of solving this NLP problem, an improved particle swarm optimization algorithm based on chaotic dynamics and adaptive inertia parameters (CAPSO) is designed. The introduction of the Logistic chaotic mapping enhances the algorithm’s global search ability, while the adaptive inertia weights improve convergence speed and solution accuracy. The experimental results demonstrate that the proposed method outperforms three traditional heuristic algorithms, achieving both higher system revenue and faster convergence.