UBITECH Ltd., 15231 Athens, Greece.
Greener than Green Technologies S.A., 14564 Athens, Greece.
Sensors (Basel). 2022 Apr 16;22(8):3068. doi: 10.3390/s22083068.
Contemporary wastewater reclamation units entail several diverse treatment and extraction processes, with a multitude of monitored quality characteristics, controlled by a variety of key operational parameters directly affecting the efficiency of treatment. The conventional optimization of this highly complex system is time- and energy- consuming, frequently relying on intuitive decision making by operators, and does not predict or forecast efficiency changes and system maintenance. In this paper, we introduce intelligent solutions to enhance the operational control of the unit with minimal human intervention and to develop an AI-powered DSS that is installed atop the sensors of a water treatment module. The DSS uses an expert model, both to assess the quality of water and to offer suggestions based on current values and future trends. More specifically, the quality of the produced water was successfully visualized, assessed and rated, based on a set of input operational variables (pH, TOC for this case), while future values of monitored sensors were forecasted. Additionally, monitoring services of the DSS were able to identify unexpected events and to generate alerts in the case of observed violation of operational limits, as well as to implement changes (automatic responses) to operational parameters so as to reestablish normal operating conditions and to avoid such events in the future. Up to now, the DSS suggestion and forecasting services have proven to be adequately accurate. Though data are still being collected from early adopters, the solution is expected to provide a complete water treatment solution that can be adopted by a vast range of parties.
当代废水回收装置涉及多种不同的处理和提取工艺，具有多种监测质量特性，由多种直接影响处理效率的关键操作参数控制。这种高度复杂系统的传统优化既耗时又耗能，经常依赖操作人员的直观决策，并且无法预测或预测效率变化和系统维护。在本文中，我们引入了智能解决方案，以最小的人为干预来增强装置的运行控制，并开发了一个基于人工智能的决策支持系统，该系统安装在水处理模块的传感器之上。该 DSS 使用专家模型，既可以评估水质，也可以根据当前值和未来趋势提供建议。更具体地说，成功地根据一组输入操作变量（在此情况下为 pH、TOC）可视化、评估和评定了所产水的质量，同时预测了监测传感器的未来值。此外，DSS 的监控服务能够识别意外事件，并在观察到违反操作限制的情况下生成警报，以及对操作参数进行更改（自动响应），以重新建立正常运行条件并避免将来发生此类事件。到目前为止，DSS 的建议和预测服务已经被证明是足够准确的。尽管仍在从早期采用者那里收集数据，但该解决方案有望提供一种完整的水处理解决方案，可供广泛的各方采用。
