Decoding the "Digital Water Tower" under the Kunlun Mountains

Multi-dimensional Stereoscopic Monitoring Perception + Twin Mapping

The Hotan Oasis in southern Xinjiang, northwest China, is embedded between the vast sea of sand and the towering mountains. The winding Karakax River flows quietly. At the exit of the middle reaches of this long river, the Uluwati Water Conservancy Project is located.

The Uluwati Water Conservancy Project plays an important role in flood control, power generation, irrigation, ecology, and sediment scheduling, protecting the safety of the downstream and the happiness of the people.

This "digital water tower" empowered by the "smart brain" under the Kunlun Mountains is better benefiting the people.

"Replica" of the digital twin platform of the physical dam

In the past, Xinjiang water conservancy workers "chased water" hard at the dam and river bank. Now, with the sensing equipment all over the dam body and the water surface along the line, the digital virtual dam displayed in three dimensions and intuitively on the "digital twin Uluwati platform" in the information hall of the Uluwati Water Conservancy Hub completely "replicated" the physical dam a few kilometers away.

Combined with the multi-dimensional three-dimensional perception of "Sky, Earth, Water and Engineering", the standardization and normalization of various data are initially realized. Based on the visualization engine, multi-source data aggregation is realized. A digital twin platform is built to provide a visual and interactive application platform for comprehensive decision-making. Focusing on the "four forecasts", breakthroughs are explored in the application of situational awareness, flood control dispatch, engineering safety, video intelligence AI, etc., to support the precise dispatch, refined management, and scientific decision-making of the project.

Situational awareness: Comprehensively integrate the existing hydrological stations, safety monitoring, reservoir water conditions, power generation, video monitoring and other monitoring data, based on the construction of the data base, multi-dimensional data governance, gather the information of water conservancy hubs and upstream and downstream elements, use the twin engine to simulate and display multi-dimensional information, realize a one-screen overview of the project, and provide data support for dispatching decisions.

Flood control dispatch Flood control dispatch forecast integrates two forecasts: the flood model forecast during the flood season and the upstream station flow inflow forecast model, which complement each other's advantages and can be handled by SMS alarm when the warning is exceeded. It can automatically make scheduling plans, compare the differences in flood discharge flow processes and reservoir water level control changes of multiple plans, select and adjust the optimal plan, and conduct simulation rehearsals to intuitively analyze and grasp the expected scheduling effects. The integrated plan function can quickly guide different response levels and deal with disposal measures and processes.

Engineering safety is based on historical and real-time monitoring data such as dam internal and external monitoring, earthquake monitoring and evaluation. The background automatically calls professional models for dam monitoring statistical regression prediction, forecasts the dam scheduling process and safety monitoring data in a certain period of time in the future, and identifies abnormal safety monitoring areas through comprehensive data and finite element analysis, helping decision-makers to grasp the future dam safety situation, assist engineering scheduling, and ensure the safe operation of hub projects.

Video AI tracks and records the trajectory of people and vehicles through thermal imaging, structured recognition, super brain edge computing and other technologies, realizes abnormal intrusion recognition, graded alarms, and improves night monitoring and intelligent security warning capabilities. Water level AI recognition is realized based on the convolutional neural network algorithm, providing multi-channel comparison and monitoring backup of key data of water conservancy hub operation.

"Showing off" in practical applications

The digital twin construction of the Uluwati Water Conservancy Hub is "showing off" in practical applications, allowing this land to say goodbye to the worries of water shortage and flood control, and constantly injecting vitality into local development

Efficiently assisting flood control scheduling decisions

On the digital twin platform, the fusion of geospatial data and multi-source data such as BIM models, real-time and historical monitoring data realizes real-time mapping of physical projects and twin projects. The three-dimensional visual scene intuitively displays the flood forecast, flood scheduling plan preview and optimization, and provides emergency response plans to efficiently assist flood control scheduling decisions. Using Beidou navigation and measurement robots for precise positioning, the millimeter-level automatic real-time monitoring of the surface deformation of the dam and slope is realized. Through multi-dimensional perception and analysis of more than 400 monitoring points inside and outside, managers can always "know the bottom line" of the project safety status. The bottom-line thinking coordinates the flood season and the construction situation, and maximizes the role of the project in flood control and water resources scheduling.

Using this technology, statistical analysis can be conducted based on the water demand of the irrigation area, water supply services can be adjusted in a timely manner, and water resources can be used "precisely". The agricultural irrigation water supply in 2024 is 1.789 billion cubic meters, exceeding the annual water supply task of 1.576 billion cubic meters. During the spring irrigation period, the reservoirs provided a total irrigation water supply of 507 million cubic meters, of which the reservoirs increased water supply by 99.14 million cubic meters, 147 million cubic meters more than the planned water supply. During the winter irrigation period, the reservoirs provided a total irrigation water supply of 134 million cubic meters, of which the reservoirs increased water supply by 34.01 million cubic meters, 48 million cubic meters more than the planned water diversion. The number of winter wheat irrigations has been increased to 5 times, providing strong guarantee services for irrigation areas to irrigate as much as possible, ensure food security and increase farmers' income.

In view of the high sediment content of the Karakax River and the serious siltation of the reservoir during the flood season, the digital twin sensing data is combined for forecasting and rehearsal, and the timing of the flood peak is selected to lower the reservoir water level, shorten the time period, and use low water level and large flow to flush and discharge sand. In 2024, 12.159 million cubic meters of sand will be discharged, and the effect of sand discharge and silt reduction is obvious, which effectively increases the reservoir's beneficial storage capacity and effectively improves the reservoir's irrigation and water supply guarantee capacity. In the next step, we will improve the front-end sensing equipment and facilities, digital twin platforms and models, intelligent business applications, and network security protection of the Uluwati Water Conservancy Hub on the basis of the existing achievements, and further enhance the multi-dimensional monitoring capabilities of the intelligent "four predictions" and management decision-making capabilities to provide strong water security guarantees for Xinjiang's high-quality development.