马从国, 赵德安, 王建国, 陈亚娟, 李亚洲. 基于无线传感器网络的水产养殖池塘溶解氧智能监控系统[J]. 农业工程学报, 2015, 31(7): 193-200. DOI: doi:10.3969/j.issn.1002-6819.2015.07.028
    引用本文: 马从国, 赵德安, 王建国, 陈亚娟, 李亚洲. 基于无线传感器网络的水产养殖池塘溶解氧智能监控系统[J]. 农业工程学报, 2015, 31(7): 193-200. DOI: doi:10.3969/j.issn.1002-6819.2015.07.028
    Ma Congguo, Zhao Dean, Wang Jianguo, Chen Yajuan, Li Yazhou. Intelligent monitoring system for aquaculture dissolved oxygen in pond based on wireless sensor network[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2015, 31(7): 193-200. DOI: doi:10.3969/j.issn.1002-6819.2015.07.028
    Citation: Ma Congguo, Zhao Dean, Wang Jianguo, Chen Yajuan, Li Yazhou. Intelligent monitoring system for aquaculture dissolved oxygen in pond based on wireless sensor network[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2015, 31(7): 193-200. DOI: doi:10.3969/j.issn.1002-6819.2015.07.028

    基于无线传感器网络的水产养殖池塘溶解氧智能监控系统

    Intelligent monitoring system for aquaculture dissolved oxygen in pond based on wireless sensor network

    • 摘要: 为了便于对规模化水产养殖池塘溶解氧的监控,该文研制了一种基于无线传感网的水产养殖池塘溶解氧智能监控系统,实现对池塘溶解氧的分布测量、智能控制和集中管理。针对常规模糊PID控制器自适应能力低,提出了一种可变论域模糊 PID 控制器,根据溶解氧误差和误差变化的大小动态调整模糊控制单元的输入输出变量论域,能较好地解决了模糊控制规则数量与溶解氧控制精度之间的矛盾,实现了PID控制器参数的自整定。根据池塘溶解氧变化的非线性、大时滞和大惯性等特点,设计基于变论域模糊PID控制器与增氧机转速PID调节器构成的池塘溶解氧串级控制系统,溶解氧控制器的输出为增氧机转速调节器的输入,增氧机转速调节器输出改变增氧机转速使溶解氧浓度快速跟踪目标值。根据溶解氧测量数值序列的变化趋势,基于灰色理论和权重构建组合灰色溶解氧预测模型,以预测值作为变论域模糊PID控制器的反馈值,实现对溶解氧的预测控制,起到超前调节的目的。在试验池塘和对照池塘分别采用变论域模糊PID控制器和模糊PID控制器对池塘溶解氧进行调控,对照池塘溶解氧的响应时间比试验池塘延长15 min左右,超调量扩大2.96倍,对照池塘溶解氧的标准差、均方差、最大误差和最小误差指标比试验池塘扩大3~4倍。试验结果表明可变论域模糊PID控制器能够改善池塘溶解氧控制系统的动态性能,提高控制系统的稳态精度,有效地抑制影响池塘溶解氧稳定的诸多不确定因素的干扰,满足水产养殖对池塘溶解氧的要求,为解决非线性和大时滞复杂对象的控制问题提供一个新的控制思路。

       

      Abstract: Abstract: In order to facilitate DO (dissolved oxygen) monitoring for a scaled aquaculture pond, a DO intelligent monitoring system was developed based on a wireless sensor network, which could realize distribution measurement, intelligent control, and centralized management. The system consists of a three-layer structure including data acquisition and control, water quality monitoring, and water management. The data acquisition and control layer was composed of data acquisition and control terminals, routing nodes, and a coordinator node based on ZigBee technology. They were deployed in the sensing area for an aquaculture pond's waters, and they constituted a wireless monitoring network for water quality environmental parameters by self organization to collect water quality parameters and adjust control devices. The water quality monitoring layer included a water quality monitoring terminal and a communication computer, which realized water quality supervision and aquaculture equipment intelligent control. The water quality management layer contained mainly a water quality management terminal, a system database, and a Web server end. The water quality management end was responsible for analysis and processing for the water quality data. The monitoring system concentrated wireless data acquisition, intelligent control, and centralized management for water quality parameters to improve scale aquaculture benefit and information management level. Aiming at low adaptive ability for conventional fuzzy PID controller, a variable universe fuzzy PID controller was proposed, which comprises an adjustment unit for expansion factors, variable universe fuzzy control unit and PID controller, the extension factor α1, α2 and β for input and output domain of fuzzy control unit were adjusted constantly by an expansion factor regulating unit according to DO error and DO error change rate. PID controller parameters were tuned online by a variable universe fuzzy control unit to realize the purpose for DO adaptive control. According to the DO change characteristics with nonlinear, large delay and large inertia in a pond, the cascade control system was constituted by a variable universe fuzzy PID controller and an aerator speed PID controller. DO was the main controlled variable, and the aerator speed was the secondary controlled variable. If the DO concentration deviates from the setting value, a DO control loop will operate. Its output is the control loop input of the aerator speed, and the output of the aerator regulator changes the aerator speed to make DO concentration fast track the setting value of system target. The cascade control system can timely and accurately adjust the DO concentration to meet aquacultural needs. According to the changing trend of the DO data sequence for multiple pond monitoring sites, a combination grey DO forecasting model was constructed based on grey theory and weights to predict DO concentration and the feedback value for a variable universe fuzzy PID controller. This achieved DO prediction control and beforehand adjustment, and DO overshooting was well restrained. The test pond and the contrast pond respectively used a variable universe fuzzy PID controller and a fuzzy PID controller to regulate DO in the ponds, and the DO target value was 7.10 mg/L. In the dynamic response phase, DO response time for the contrast pond extended about 15 min more than the test pond, and the overshooting expanded 2.96 times. After the system entered steady process, the standard deviation, the mean variance, and the maximum error and minimum error enlarged by 3-4 times. The testing results showed that the adjustment process for the test pond had characteristics with fast response, small overshooting, high control precision, and good stability compared with the contrast pond. The adjusting factors for the variable universe fuzzy PID controller better solved contradictions between the quantity of fuzzy control rules and DO control precision, and realized self tuning for PID parameters, and improved dynamic performance, and raised steady accuracy and quality of the fuzzy PID controller. It can effectively restrain many uncertain factors of interference affecting DO stability to meet aquaculture requirements for DO and provide a new control method to solve control problem for complex objects with a nonlinear and large time delay.

       

    /

    返回文章
    返回