Abstract: In order to accurately establish the tripping time prediction model of circuit breaker thermal tripping device and improve the stability of the tripping device, an Improved Whale Optimization Algorithm（IWOA） and Support Vector Regression（SVR） modeling method are proposed. In view of the problems of Whale Optimization Algorithm（WOA）, such as easy to get into local optimal solution, slow convergence speed and low convergence accuracy, three strategy improvements are proposed and compared on ten basic test functions. The results prove that the improved algorithm has higher convergence accuracy and faster convergence speed. Then, the prediction model of iwoa-svr tripping time is established, and the production process of circuit breakers is optimized, so as to improve the production efficiency and product quality of circuit breakers.

Key words: tripping time, comprehensive modeling, whale optimization algorithm, production optimization

摘要： 为了准确地建立断路器热脱扣器的脱扣时间预测模型，改善脱扣器的稳定性。提出了一种改进的鲸鱼优化算法（IWOA）和支持向量回归（SVR）综合建模的方法。针对鲸鱼优化算法（WOA）具有易陷入局部最优解、收敛速度慢、收敛精度低等问题，提出三个策略改进，并在10个基本测试函数上进行比较，结果证明改进的鲸鱼优化算法具有更高的收敛精度和更快的收敛速度。建立IWOA-SVR脱扣时间预测模型，并优化断路器的生产过程，进而提高断路器的生产效率和产品质量。

关键词: 脱扣时间, 综合建模, 鲸鱼优化算法, 生产优化