Reliability Optimization Of Central Heating System Based On Particle Swarm Optimization Algorithm

The heat provisions needed in many cities of China are increasing day by day and the scale of central heating is becoming larger. Therefore reliabilities of heating networks are required much higher. However, the design stages of heating systems in China usually lack the considerations of reliability; the operation and management abilities of some heating networks are relatively low, which cause the high failure rates of heating networks. It is clear that we should be conscious of the rule of heating network failures and enhance the reliability at the expenses of economic performance. Thus, it is necessary to implement a reliability optimization study of heating networks, considering reliability and economic performance simultaneously.Through the investigation of typical heating networks in Jilin and Changchun, failure data of heating pipelines, valves and compensators were collected and failure rate statistics and spectrum analyses were carried out subsequently based on these data. Furthermore, a mathematical reliability optimization model with initial capital cost of heating network as objective function and reliability criterion as the primary constraint was established in this paper. Later, this model combined with particle swarm optimization (PSO) algorithm based on random inertia weight (RIW) strategy was successfully applied to the Yuanyuan heating network of Jilin. The reliability criterion and initial capital cost after optimization is0.9044and6.433x10’Yuan, which increase7.83%and8.81%respectively compared with those before optimization; the investment of accessories account for about84.3%of the total increment of the initial capital investment. Additionally, it can be inferred from the optimized heating network that one of the best way to increase reliability is setting valves in heating branches with relatively high heat load. Namely, the more heat load sectioning valve or cut-off valve control, the higher reliability is. Afterwards, reliability-economy performance criterion ER was introduced in accordance with the engineering economics. ER represents the ratio of reliability criterion and initial capital cost of a heating network. For Yuanyuan heating network, ER is rather low without taking any reliability-increasing measures, the value is0.0578, but ERs are larger than0.1for any other situations with reliability-increasing measures; ER will decline steadily as long as the reliability criterion increases, especially if reliability reaches a certain value, ER will drop down dramatically, which means it is very hard to improve the reliability criterion any further.The particle swarm optimization is introduced in this paper for the first time to realize the reliability optimization of heating networks using MATLAB scientific computing software. At the same time, the reliability-economy performance criterion ER is also brought forward. They are of importance to some extent as guidance for enhancing the reliabilities of heating networks.