中文摘要：

流程工业具有连续性，实时性，复杂性等特点，其生产调度的优化一直是行业内的热点问题。利用蚁群算法进行了深入的研究，提出了适用于流程工业中的连续域蚁群算法方案。该算法采用连续域网格对流程工业问题进行建模，利用缩小范围重新划分的方法提高了算法精度，同时引入最大一最小蚁群算法保证收敛速度，并通过限制解的范围和估算产值的方法引导蚂蚁走向可行，高效的路线。最后通过隔膜烧碱车间的算例验证了该方法在实用性、精度及速度方面有着明显的优势。

英文摘要：

The process industry covers many industries, it has the characteristics of continuity, timeliness, and complexity, so the process industry optimization is a hot issue in industry. Ant colony algorithm（ACO） is highly accurate, fast and flexible to implement. But due to lack of appropriate coding method and the guidance of heuristic algorithm, ACO has not been applied to the process industry scheduling problem before. In order to take advantage of ACO to solve the problem, this paper proposed a continuous domain ACO map system based on Uniform time Discretization Model （UDM）, and the significance of each node can be expressed through the proaess~ng capability. Most of the illegal solution could be avoided by real-time monitoring of the storage status and production capacity, and the illegal solution couldn＇t be avoided would be punished either. For the difficulties to get the output directly during the calculation, intimidate results are estimated by intimidate products to evaluate the solution. Meanwhile, MAX-MIN ACO and 2-way convergence ACO were introduced to reduce the possibility of appearing local optimum and to accelerate the convergence speed. Finally, in order to examine the practicabilility of ACD, the calculation of the process of diaphragm caustic soda shop with ACO was conducted. It shows that solving the process industry problem by ACO is practical and profitable.