生猪养殖业污水排放智慧监管系统的设计与实现

为了对生猪养殖业污水的治理过程进行监控和违规排污预警,该文提出了养殖污水实时监管策略,设计并实现了生猪养殖污水治理智慧监管系统。该系统通过信息采集模块收集养殖污水排放的实时数据,实现养殖污水实时数据监测、预警分析等功能。其中集中治理的监管是根据安装在槽罐车上的GPS数据和污水集中处理厂的信息,判断污水是否被运送到指定地点排放;工业治理的监管是采用模糊推理理论,以监管因子的浓度偏差及偏差变化率为输入量,相应的污水预警等级作为输出量,对监管因子进行模糊化及逻辑推理,建立相应的模糊监管子系统,生成工业治理监管规则及策略;针对生态治理的监管,构建了相应的监管策略和Ecological数学模型,该模型以监管策略为依据,对实时数据进行定性与定量分析预测,实现对偷排漏排、满溢等违规排污现象的判断。试验结果表明,系统预警准确度为96.17%,平均误差时间为33.22 s,违规排污量平均值为15.77 L,能够满足养殖污水排放监管要求,对提高监管效率具有重要意义。

Design and implementation of smart supervisory system for sewage emission in swine breeding industry

Abstract: Swine effluent poses not only serious threats to the environment and soil but also affect the sustainable development of livestock industry. So the treatment and supervision of swine sewage is imperative. In this paper, we divided the swine treatment methods into ecological treatment, industrial treatment and centralized treatment. To realize the remote monitoring and illegal discharge warning of swine effluent disposed by the three treatment methods, a smart supervisory system and real time data supervisory strategy was designed and proposed. Based on Internet, the newly developed system collected and monitored the effluent information such as rainfall, level, flow, location, chemical oxygen demand and biochemical oxygen demand in water, ammonia concentration and so on through data acquisition system. The system was mainly comprised of sensors, microprocessors, and GPRS wireless transmission module. The collected data were transmitted to the cloud server as real-time data for the entire supervisory system. According to the change rules of data and the practical demand of supervision, real-time data was obtained every 10 minutes. In addition, to determine the credit rating of the industries and push the alarm information in different conditions, the system would select the principle identity as the primary index of warning decision, while for the real-time warning, industry credit rating, alarm grade, and alarm processing time as the secondary indexes were used. Furthermore, the index system of warning-push was proposed in this paper. Combing the real time data and index system of warning-push, the system could realize functions of real time data monitoring, video monitoring, site management, equipment management, data statistic and early warning management. A lower inter-module coupling and a higher cohesion between the functions were designed to improve its scalability and flexibility, and to ensure that the modules could be interrelated in the logical structure and independent in the physical structure. In the centralized treatment, the system used the GPS data and basic information of sewage disposal plant to judge whether the tank car took the sewage to the designated place to discharge it or not. Based on fuzzy reasoning theory, the supervisory system of industrial treatment with five monitoring subsystems, which had two inputs and one output variables, was established. In this fuzzy supervisory system, suspended solid difference, ammonia-nitrogen difference, biochemical oxygen demand difference, chemical oxygen demand difference, total phosphorus difference and its variation rate were selected as input variables, and corresponding warning rank as an output variables. Also, symmetry triangle was selected as the membership function of input and output variables to ensure the sensitivity and robustness of the system. According to the effluent discharge standards of the livestock and poultry industry, fuzzy reasoning rules of the subsystem can be generated. To monitor the ecological treatment of swine effluent, in this paper, we constructed the corresponding supervision strategy and ecological mathematical model. Based on the principle of the supervisory strategy, the model predicted and analyzed the real-time data by qualitative and quantitative method to calculate the discharging interval during rainy days and to judge whether there existed the phenomenon of night discharge, stealing and leaking, overflowing and so on. Through the simulation test, the results showed that the system''s average warning accuracy reached 96.17%, the average error time was 33.22 s, the illegal discharge capacity caused by error time was 15.77 L. Therefore, this supervisory system could meet the requirements of swine effluent discharge regulation and improve the supervision efficiency significantly.