In the context of Industry 4.0 and digital transformation, information security has emerged as a critical challenge. Consortium blockchain, with its semi-open and multi-institution co-governance characteristics, has become a core technology for cross-organizational trusted collaboration. However, existing consensus algorithms face bottlenecks in performance and security: the high communication complexity of Practical Byzantine Fault Tolerance (PBFT) makes it difficult to meet the high-throughput demands of industrial scenarios, while Hotstuff lacks a punishment mechanism for Byzantine nodes and is vulnerable to predictable primary node attacks. This paper proposes an efficient consensus algorithm based on a reputation scoring model and Verifiable Random Function (VRF)—E-Hotstuff. By designing a dynamic reputation model to quantify node credibility, selecting a small number of high-reputation nodes to form a consensus committee, and using VRF to randomly choose the primary node, E-Hotstuff effectively suppresses the malicious behavior of Byzantine nodes and further increases transaction throughput and reduces latency. Experiments show that under a node scale of 97, the consensus throughput of E-Hotstuff is 13.46% higher than that of R-PBFT, and the latency is reduced by 13.18%. Moreover, its performance significantly outperforms R-PBFT and Hotstuff as the number of nodes increases. This paper provides an innovative solution for efficient and secure consensus in consortium blockchains for industrial scenarios.