Winter ice damage prevention method and system for long distance water delivery channel

Abstract

The application discloses a winter long-distance water conveying channel ice disaster prevention method and system, relates to the technical field of channel ice disaster prevention, and comprises the following steps: acquiring multi-modal data along a long-distance channel of a water conveying channel at a winter time of the water conveying channel, determining multi-scale regulation strategies between different channel sections according to the recognition of deduction results; simultaneously extracting water level data and water temperature data of the long-distance channel at the current time, and determining a target flow variation curve and a target water temperature control interval in combination with an ice disaster prevention and control mechanism of the long-distance channel; triggering linkage adjustment of gates of the upstream and the long-distance channel based on a dynamic cooperative control instruction, collecting ice condition images and hydraulic element data along the long-distance channel, determining deviation content between actual ice condition evolution trends and deduction results in combination with a preset deviation evaluation mechanism, and determining a plurality of ice disaster prevention projects according to the tracing of the deviation content, so that the accuracy of an ice disaster prevention sequence of the water conveying channel is improved.

Description

Technical Field

[0001] This invention relates to the technical field of ice damage prevention in waterways, and particularly to a method and system for preventing ice damage in long-distance water conveyance channels during winter. Background Technology

[0002] The uneven spatial and temporal distribution of water resources in northern my country makes long-distance inter-basin water transfer projects a key measure to address water shortages. However, during winter operation, water conveyance channels face severe ice damage threats due to exposure to low temperatures. Ice damage in winter water conveyance channels mainly manifests as ice accumulation, ice cover formation and evolution, ice jams and ice dams, posing significant challenges to the safe operation of the project and the guarantee of water supply.

[0003] In the actual operation of ice damage prevention, existing technologies often operate hydraulic scheduling and thermal compensation separately. Gate regulation mainly relies on water level and flow safety, while heating facilities mostly adopt simple constant temperature control or timer control. There is a lack of coordinated linkage between the two based on the ice damage prevention and control mechanism. This results in the inability to match a good ice prevention effect when facing complex ice conditions. For example, simply increasing the flow rate may exacerbate the formation of ice crystals due to excessively low water temperature, and simply heating the water body may cause local ice damage to still exist due to uneven flow velocity distribution. This affects the accuracy of coordinated control commands along long-distance channels, resulting in low accuracy of the ice damage prevention sequence in water conveyance channels. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art. This invention provides a method and system for preventing ice damage in long-distance water conveyance channels during winter.

[0005] This invention provides a method for preventing ice damage in long-distance water conveyance channels during winter, including:

[0006] During the winter season, multimodal data along the long-distance water conveyance channel is acquired to extrapolate the spatiotemporal evolution of the heat exchange process and ice cover formation process along the long-distance channel, thereby outputting the extrapolation results. Based on the identification of the extrapolation results, multi-scale control strategies between different channel sections are determined.

[0007] In response to the multi-scale control strategy, the system simultaneously extracts water level and temperature data along the long-distance channel at the current moment, and combines the ice damage prevention and control mechanism along the long-distance channel to determine the target flow change curve and the target water temperature control range, thereby generating collaborative control instructions along the long-distance channel.

[0008] Based on the dynamic collaborative control command, the gates along the upstream and long-distance channels are linked and adjusted, and multiple fixed-point disturbance devices are started simultaneously to break local ice crystals by frequency conversion disturbance. At the same time, the output power of the heating facilities is adjusted to compensate for the heat loss of the water.

[0009] Ice condition images and hydraulic element data are collected along long-distance channels. Combined with a pre-set deviation assessment mechanism, the deviation between the actual ice condition evolution trend and the projection results is determined. Based on the traceability of the deviation, multiple ice damage prevention and control projects are determined, thereby determining the ice damage prevention and control sequence for the water conveyance channel.

[0010] This invention provides a system for preventing ice damage in long-distance water conveyance channels during winter. This system is applied to the aforementioned method for preventing ice damage in long-distance water conveyance channels during winter. The system includes:

[0011] The water conveyance channel module is used to acquire multimodal data along the long-distance water conveyance channel during winter, to extrapolate the spatiotemporal evolution of the heat exchange process and ice cover formation process along the long-distance channel, and to output the extrapolation results. Based on the identification of the extrapolation results, multi-scale control strategies between different channel sections are determined.

[0012] The collaborative control module is used to respond to multi-scale regulation strategies, while extracting water level and water temperature data along the long-distance channel at the current moment, and combining the ice damage prevention and control mechanism along the long-distance channel to determine the target flow change curve and the target water temperature control range, thereby generating collaborative control instructions along the long-distance channel.

[0013] The linkage adjustment module is used to trigger the linkage adjustment of gates along the upstream and long-distance channels based on dynamic collaborative control commands, and simultaneously start multiple fixed-point disturbance devices to break local ice crystals by frequency conversion disturbance, while adjusting the output power of heating facilities to compensate for the heat loss of the water.

[0014] The ice damage prevention sequence module is used to collect ice condition images and hydraulic element data along long-distance channels. Combined with a preset deviation assessment mechanism, it determines the deviation between the actual ice condition evolution trend and the projection results. Based on the traceability of the deviation, multiple ice damage prevention projects are determined, thereby determining the ice damage prevention sequence of the water conveyance channel.

[0015] Compared with the prior art, the beneficial effects of the present invention are:

[0016] (1) During the winter of the water conveyance channel, multimodal data along the long-distance channel are acquired, and the spatiotemporal evolution of the heat exchange process and ice cover formation process along the long-distance channel is extrapolated. The extrapolation results are then output, and the multi-scale control strategy between different channel sections is determined based on the identification of the extrapolation results. In response to the multi-scale control strategy, the water level and water temperature data along the long-distance channel at the current moment are extracted, and the target flow change curve and target water temperature control range are determined in combination with the ice damage prevention and control mechanism along the long-distance channel. This generates a collaborative control command along the long-distance channel. The extrapolation results are introduced to control the multi-scale control strategy between different channel sections, which improves the accuracy of the collaborative control command along the long-distance channel.

[0017] (2) Based on the dynamic collaborative control command, the gates along the upstream and long-distance channels are linked and adjusted, and multiple fixed-point disturbance devices are started simultaneously to destroy local ice crystals by frequency conversion disturbance. At the same time, the output power of the heating facilities is adjusted to compensate for the heat loss of the water body. Ice condition images and hydraulic element data along the long-distance channels are collected, and the deviation between the actual ice condition evolution trend and the simulation result is determined by the preset deviation assessment mechanism. Multiple ice damage prevention and control projects are determined based on the traceability of the deviation content, thereby determining the ice damage prevention and control sequence of the water conveyance channel. The fixed-point disturbance devices and heating facilities are further controlled, and the deviation between the actual ice condition evolution trend and the simulation result is fully considered, which improves the accuracy of the ice damage prevention and control sequence of the water conveyance channel. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the method for preventing ice damage in long-distance water conveyance channels during winter, as described in this embodiment of the invention.

[0019] Figure 2 This is a schematic diagram of the structural composition of the winter long-distance water conveyance channel ice damage prevention system in an embodiment of the present invention.

[0020] Explanation of reference numerals in the attached figures:

[0021] 21. Water conveyance channel module; 22. Collaborative control module; 23. Linkage regulation module; 24. Ice damage prevention sequence module. Detailed Implementation

[0022] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0023] Please see Figure 1 A method for preventing ice damage in long-distance water conveyance channels during winter, applied to scenarios involving ice damage in water conveyance channels; the method includes:

[0024] Step S11: During the winter season of the water conveyance channel, acquire multimodal data along the long-distance channel, perform spatiotemporal evolution simulation of the heat exchange process and ice cover formation process along the long-distance channel, and output the simulation results. Based on the identification of the simulation results, determine the multi-scale control strategy between different channel sections.

[0025] Step S12: In response to the multi-scale control strategy, the water level and water temperature data along the long-distance channel at the current moment are extracted, and the target flow change curve and target water temperature control range are determined by combining the ice damage prevention and control mechanism along the long-distance channel, thereby generating a collaborative control command along the long-distance channel.

[0026] Step S13: Based on the dynamic collaborative control command, trigger the linkage adjustment of the gates along the upstream and long-distance channels, simultaneously start multiple fixed-point disturbance devices to break local ice crystals in the form of frequency conversion disturbance, and adjust the output power of the heating facilities to compensate for the heat loss of the water.

[0027] Step S14: Collect ice condition images and hydraulic element data along the long-distance channel, combine them with the preset deviation assessment mechanism to determine the deviation between the actual ice condition evolution trend and the projection results, and determine multiple ice damage prevention and control projects based on the traceability of the deviation content, thereby determining the ice damage prevention and control sequence of the water conveyance channel.

[0028] In step S11, the specific steps are as follows:

[0029] S111: For water conveyance channels during winter, collect multimodal data along long-distance channels. The multimodal data includes meteorological field data, water temperature distribution data, flow velocity field data, and ice cover thickness point cloud data. Dynamically extrapolate the multimodal data and perform real-time inversion of the heat exchange process along long-distance channels. In addition, combine ice-hydraulic dynamics to extrapolate the spatiotemporal evolution of the ice cover formation process, and output the extrapolation results, which include the spatiotemporal distribution characteristics of ice condition evolution.

[0030] S112: Based on the identification of the simulation results, the corresponding ice flower density and ice flow volume are determined, and the corresponding high-risk area, transitional stable area and ice-free water conveyance area are determined by combining the distribution images along the long-distance channel. Based on the thermal imbalance state of the high-risk area, transitional stable area and ice-free water conveyance area, the multi-scale control strategy between different channel sections is determined. The multi-scale control strategy includes the flow pulse strategy on an hourly scale and the heat storage and insulation strategy on a daily scale.

[0031] In the embodiments of this application, multimodal data along a long-distance channel is collected during winter. The multimodal data includes meteorological field data, water temperature distribution data, flow velocity field data, and ice cover thickness point cloud data. The multimodal data is dynamically extrapolated, and the heat exchange process along the long-distance channel is inverted in real time. Furthermore, the spatiotemporal evolution of the ice cover formation process is extrapolated by combining ice-water dynamics, thereby outputting the extrapolation results. The extrapolation results include the spatiotemporal distribution characteristics of ice condition evolution and are introduced into the extrapolation results.

[0032] At this time, for the winter operation of the water conveyance channel, a three-dimensional sensing network covering the long-distance channel is constructed to simultaneously collect multimodal data, including meteorological field data, water temperature distribution data, flow velocity field data, and ice cover thickness point cloud data. During the collection process, the meteorological field data is sampled at high frequency to obtain the temperature, wind speed, wind direction, and solar radiation flux data along the route, which are used as boundary conditions for calculating the heat flux at the atmosphere-water interface.

[0033] Vertical stratification of water temperature distribution data is performed. Temperature sensor arrays deployed at different depths in the channel are used to obtain vertical water temperature gradient data from the water surface to the bottom layer, which is used to analyze the thermal stratification of the water body. Three-dimensional flow field scanning is performed on the flow velocity field data to obtain the flow velocity vector distribution in the channel cross section, focusing on monitoring the vertical variation law of flow velocity and the intensity of lateral circulation.

[0034] Non-contact scanning of ice cover thickness point cloud data is performed using mobile inspection equipment equipped with lidar or structured light cameras to obtain high-density three-dimensional point cloud coordinates covering the channel surface. After acquiring the above multi-source data, a unified time benchmark and spatial grid reference system are established. Multimodal data of different frequencies and accuracies are spatiotemporally aligned and interpolated to form a spatiotemporally consistent multimodal dataset, providing a data foundation for subsequent dynamic simulation.

[0035] A channel unsteady-state thermodynamic inversion model is constructed based on a multimodal dataset to perform real-time calculations of the heat exchange process along long-distance channels. In this process, the sensible heat flux and latent heat flux at the water-air interface are calculated based on the air temperature, wind speed and radiation in the meteorological field data, combined with the surface water temperature in the water temperature distribution data.

[0036] Based on the ice edge location and ice cover rate extracted from the ice cover thickness point cloud data, the open channel section and the ice-sealed section are distinguished. The latent heat flux of phase change at the water-ice interface is calculated for different channel sections. This flux characterizes the heat absorption or release intensity of the water body during the phase change process. At the same time, combined with the geological parameters of the channel bed and water temperature distribution data, the heat conduction flux between the water body and the channel bed soil is calculated. By solving the water body energy balance equation simultaneously, the heat fluxes are superimposed and integrated to obtain the net heat exchange flux distribution along the channel at the current moment. Based on this, the key nodes of thermal imbalance, namely the areas where the water body heat loss rate is higher than the average level, are identified, thereby quantifying the dynamic change characteristics of the heat exchange process.

[0037] At this point, the sensible heat flux at the water-air interface is equal to the product of air density, specific heat at constant pressure of air, exchange coefficient, wind speed, and the difference between the water surface and air temperature; the latent heat flux is calculated using Dalton's evaporation formula, which is equal to the product of the latent heat of vaporization, evaporation coefficient, wind speed, and the difference between the saturated specific humidity of the water surface and the specific humidity of the air.

[0038] The latent heat flux of phase change at the water-ice interface is calculated based on the ice cover rate and the average water temperature beneath the ice. When the water temperature is below zero degrees Celsius, the water releases latent heat to form ice crystals, and the release rate is equal to the ice crystal formation rate multiplied by the latent heat of water's freezing. The heat conduction flux between the water body and the canal bed is calculated using the one-dimensional heat conduction equation. This flux is equal to the thermal conductivity of the canal bed soil multiplied by the difference between the water temperature and the canal bed temperature, and then divided by the thickness of the thermal boundary layer. The sum of the above heat fluxes is divided by the product of the water volume and specific heat to obtain the rate of change of water temperature over time. By integrating second by second using the forward Euler method, the dynamic evolution of the water temperature along the canal can be inverted.

[0039] Based on the fundamental principles of ice-hydraulic mechanics and thermodynamic inversion, the spatiotemporal evolution of ice cover formation is extrapolated. The water body is considered a two-phase flow system composed of liquid water and solid ice (including ice flowers, ice discs, and ice caps), and a water-ice coupled dynamic equation is established. The supercooling of the water body is calculated based on the thermodynamic inversion results. When the water temperature is below the phase transition critical point, an ice crystal growth model is activated to simulate the nucleation and buoyancy of ice crystals in the water. Velocity field data is used to drive a flowing ice transport model to simulate the drift trajectory of ice crystals and ice discs within channels, focusing on calculating the collision probability between ice particles, the bonding efficiency, and the probability of floating ice accumulation at obstacles or narrow points.

[0040] An ice sheet thickness evolution equation is introduced, and the increase and decrease of ice sheet thickness are calculated based on the heat exchange intensity and water flow shear force at the bottom of the ice sheet. Through spatiotemporal integration, the dynamic process of the ice sheet front advancing from downstream to upstream, as well as the thermodynamic-mechanical state evolution inside the ice sheet, are deduced. The deduction results, which include the spatiotemporal distribution characteristics of ice condition evolution, are output. These results cover core parameters such as ice sheet formation location, advancing velocity, ice sheet thickness distribution, and subglacial flow field structure.

[0041] Specifically, in the bend section of the S-shaped channel, meteorological data show that local wind speeds are reduced due to topographic obstruction, but solar radiation exhibits significant spatiotemporal nonuniformity due to the shading of the two bank slopes, i.e., the convex bank has shorter sunshine duration and weaker radiation, while the concave bank has longer sunshine duration and stronger radiation. Based on this, the collected water temperature distribution data reflects that the surface water temperature on the convex bank side is significantly lower than that on the concave bank side. At this time, the ice cover thickness point cloud data collected by lidar reveals that due to the lower flow velocity on the convex bank, according to the flow velocity field data, the circulation in the bend causes the flow velocity on the convex bank to slow down, and there is a backflow zone, where floating ice is prone to stagnation and accumulation.

[0042] Based on the above data, a real-time inversion of the heat exchange process was performed: the calculation revealed that the heat flux at the water-air interface in the convex bank region exhibited a strong heat loss state, and due to the accumulation of ice flowers leading to an increase in the thermal resistance at the water-ice interface, the sensible heat of the water could not be effectively released, further exacerbating the local supercooling of the water and causing a severe thermal imbalance in the region.

[0043] Based on the principles of ice-water dynamics, the model simulation shows that under the influence of centrifugal force in the bend, surface ice flows towards the convex bank, colliding and adhering with static floating ice trapped there at high frequency. The simulation results indicate that the inner side of the convex bank of the S-shaped channel will become the starting point for ice cover formation, which will gradually extend towards the center of the channel and upstream. The simulation results clearly indicate that within the next 4 hours, a nascent ice cover with a thickness of about 5-10 cm will form on the convex bank of the S-shaped bend, with irregular serrated edges. This spatiotemporal distribution characteristic of ice condition evolution directly predicts the risk of ice blockage at this location.

[0044] Furthermore, based on the identification of the simulation results, the corresponding ice flower density and ice flow volume are determined. Combined with the distribution images along the long-distance channel, the corresponding high-risk area for ice damage, transitional stable area, and ice-free water conveyance area are identified. Based on the thermal imbalance state of the high-risk area for ice damage, transitional stable area, and ice-free water conveyance area, multi-scale control strategies are determined between different channel sections. The multi-scale control strategies include flow pulse strategies on an hourly scale and heat storage and insulation strategies on a daily scale, ensuring the accuracy of the multi-scale control strategies.

[0045] At this point, based on the simulation results output in step S111, the spatiotemporal distribution characteristics of ice condition evolution are analyzed, and key hydraulic-thermal characteristic parameters are extracted using a gridded statistical method. In this process, for the ice phase distribution field in the simulation results, the proportion of ice flake volume in a unit volume of water is calculated and defined as ice flake density. At the same time, combined with the flux integral of the flow direction vector in the velocity field data, the total volume of ice passing through a specific cross section per unit time is calculated and defined as ice flow volume. By continuously scanning the ice flake density and ice flow volume at each calculation node along the long-distance channel, an ice condition intensity distribution curve along the channel is established. Channel sections with ice flake density greater than the first preset threshold and ice flow volume showing a monotonically increasing trend are identified as potential ice hazard outbreak points. Thus, the qualitative spatiotemporal distribution of the simulation is transformed into quantifiable numerical indicators, providing data support for subsequent regional division.

[0046] By combining distribution images along long-distance channels, including visible light monitoring images, infrared thermal imaging images, and satellite remote sensing images, spatial verification and correction are performed on the identification results, thereby dynamically dividing long-distance channels into three types of control areas. The specific division logic is as follows: areas with extremely high ice flower density, fast ice accumulation rate, and obvious ice bridge or ice cover leading edge features in the distribution images are defined as high-risk areas for ice damage.

[0047] The region with moderate ice flake density, fully formed or dynamically balanced ice cover, and relatively uniform low temperature characteristics in infrared images is defined as the transitional stable region; the region with ice flake density approaching zero, water temperature consistently above the phase transition point, and open flow in images is defined as the ice-free water transport region; during the division process, the topological relationship between regions is established, the boundary nodes and spatial span of each region are determined, and a full-line ice condition map is formed, which serves as the spatial basis for differentiated regulation.

[0048] For the identified high-risk ice-damage zones, transitional stable zones, and ice-free water conveyance zones, their corresponding thermal imbalance states are calculated, i.e., the difference between the actual heat flux of the water body and the heat flux required to maintain the current ice state. Based on the analysis results of the thermal imbalance states, a pre-set control strategy library is invoked to determine multi-scale control strategies for different canal sections. These multi-scale control strategies include two dimensions on the time scale:

[0049] For high-risk areas of ice damage, when rapid thermal loss is detected and the risk of ice blockage is imminent, a flow pulse strategy on an hourly scale is generated, which aims to disrupt the critical conditions for ice cover formation by changing hydraulic conditions in a short period of time.

[0050] For transitional stable zones and ice-free water conveyance zones, when long-term meteorological changes are predicted to lead to insufficient heat reserves, a heat storage and insulation strategy on a daily scale is generated. This strategy aims to maintain the thermal stability of the channel by optimizing the use of sensible heat from the water body and ambient thermal energy. It also aims to align and resolve conflicts between strategies in different spatial regions in time and space, and output a long-distance channel multi-scale control scheme that includes control objectives, execution time, and target objects.

[0051] Optionally, after obtaining the projection results containing the spatiotemporal distribution characteristics of ice condition evolution, the ice flake density and ice flow volume in each discrete calculation unit are first quantitatively extracted. The ice flake density is calculated as follows: in the water volume of the unit, the ice phase volume fraction value output by the ice-water two-phase flow model is used to divide the ice flake volume by the total water volume to obtain the volume percentage, expressed as a percentage.

[0052] The method for calculating the volume of flowing ice is as follows: take a cross-section of unit width perpendicular to the direction of water flow, multiply the flow velocity of each layer on the cross-section by the density of ice flowers in the corresponding layer, integrate along the water depth, and then multiply by the width of the cross-section to obtain the total volume of ice passing through the cross-section per unit time, in cubic meters per minute.

[0053] Based on the two quantitative indicators mentioned above, this invention sets the following three preset thresholds as the basis for area division: the first preset threshold is an ice flower density greater than or equal to 0.3%, the second preset threshold is an ice flower density between 0.1% and 0.3%, and the third preset threshold is an ice flower density less than 0.1%. Simultaneously, considering the changing trend of ice flow volume, when the ice flow volume shows a monotonically increasing trend along the route and the increase per kilometer exceeds five cubic meters per minute, the determination of high-risk areas for ice damage is further strengthened.

[0054] The specific logic for regional division is as follows: Channel sections that simultaneously meet the conditions of ice flower density greater than or equal to 0.3%, monotonically increasing ice flow volume, and maintaining this state within three or more consecutive calculation units, combined with areas where obvious ice bridges or white continuous ice cap leading-edge features can be observed in the distribution image, are defined as high-risk ice and snow zones; areas with ice flower density between 0.1% and 0.3%, where ice caps have formed but the thickness change rate is within ±1 cm per hour, and infrared images show uniform temperature distribution, are defined as transitional stable zones; areas with ice flower density less than 0.1%, water temperature consistently above 0.5 degrees Celsius, and exhibiting continuous open flow in visible light images are defined as ice-free water transport zones.

[0055] Specifically, based on the simulation results of the S-shaped bend convex bank, the system identified that the ice flower density in this area at the current moment has reached 0.35%, exceeding the preset first threshold of 0.3%, and the ice flow rate remains at 15m. 3The high level of / min; combined with the visible light monitoring images along the S-shaped channel, it can be clearly observed that a large amount of ice is swirling and accumulating on the inner side of the convex bank of the bend, and the ice flakes have begun to connect with each other to form loose ice bridges. Therefore, the convex bank section of the S-shaped bend and its downstream 200-meter range are clearly defined as a high-risk area for ice damage.

[0056] For the smooth transition section upstream of the S-shaped curve, the simulation results show that the ice sheet thickness is stable at 5 mm. At a depth of 10 cm, infrared thermal imaging showed that the water temperature under the ice remained at around 0.02 degrees Celsius, and the heat exchange was in a relatively balanced state. Therefore, this area was classified as a transitional stable zone. However, in the upstream rapid slope section of the S-shaped channel, due to the high flow velocity and strong aeration, the water temperature remained above 0.5 degrees Celsius, and the image showed no ice crystals. Therefore, it was classified as an ice-free water conveyance zone.

[0057] Based on the thermal imbalance state of the above-mentioned regions, a multi-scale control strategy is formulated:

[0058] For the "high-risk area for ice damage: S-bend convex bank", due to the rapid heat loss and insufficient hydraulic ice transport capacity in this area, the system triggers a flow pulse strategy on an hourly scale. The specific instruction is: in the next 11 hours, by adjusting the upstream gate, a flow pulse wave with an amplitude of 10% and a duration of 30 minutes is generated. The water surface fluctuation and shear force increment generated by the pulse are used to destroy the ice bridge structure that is forming on the convex bank, and the accumulated ice flowers are forcibly washed and transported to the downstream wide and shallow channel section, thereby mitigating the risk of ice blockage.

[0059] For the "transitional stable zone: upstream straight section", although the current situation is stable, the temperature is predicted to drop sharply by 8 degrees Celsius in the next 48 hours, posing a risk of overall heat deficit. The system has formulated a heat storage and insulation strategy on a daily scale. The specific instructions are: within the next 24 hours, the operating water level of the entire canal section will be raised by 0.5 meters to increase the heat capacity of the water body and enhance its ability to resist cold waves. At the same time, the higher water depth will reduce the heat dissipation area of ​​the water surface to ensure that the water temperature at the S-shaped bend does not drop below freezing point, thereby ensuring the water conveyance safety of the S-shaped canal in all time dimensions.

[0060] In step S12, the specific steps are as follows:

[0061] S121: In response to the multi-scale control strategy, the system extracts the current water level and water temperature data from the distributed sensor array along the long-distance channel, inputs the water level and water temperature data into the preset collaborative decision model, and performs multi-objective optimization by combining the ice damage prevention and control mechanism along the long-distance channel and the corresponding constraint relationship, thereby determining the target flow change curve to suppress ice cover formation and the target water temperature control range to maintain stable water delivery under the ice cover.

[0062] S122: Cross-match the target flow rate change curve with the target water temperature control range, and decouple the spatiotemporal coupling relationship between hydraulic regulation and thermal compensation during the matching process. Analyze the spatiotemporal coupling relationship, and determine the coordinated control instructions along the long-distance channel by combining the current status along the channel during the analysis process. The coordinated control instructions include the timing allocation instructions for upstream reservoir discharge and the operating parameters of each control gate along the channel.

[0063] In the embodiments of this application, in response to the multi-scale control strategy, the current water level data and water temperature data are extracted from the distributed sensor array along the long-distance channel. The water level data and water temperature data are input into the preset collaborative decision model, and multi-objective optimization is performed in combination with the ice damage prevention and control mechanism and corresponding constraint relationship along the long-distance channel. This determines the target flow change curve to suppress ice cover formation and the target water temperature control range to maintain stable water delivery under the ice cover, thus incorporating the overall consideration of the ice damage prevention and control mechanism and corresponding constraint relationship along the long-distance channel.

[0064] At this point, in response to the multi-scale control strategy determined in step S112, the system immediately initiates the data preprocessing process of the collaborative control command; during this process, the water level and water temperature data at the current moment are extracted with millisecond-level time synchronization accuracy through the distributed sensor array deployed along the long-distance channel.

[0065] At this point, real-time water depth and elevation data for each key node are obtained from ultrasonic and pressure level gauges deployed on the channel cross-section to construct a longitudinal profile of the water surface along the channel. Simultaneously, real-time water temperature data at different vertical depths are extracted from high-precision resistance sensors integrated into the water temperature monitoring chain to construct a water temperature distribution field along the channel. To eliminate data jitter caused by sensor random noise and environmental interference, the extracted water level and water temperature data are subjected to time window moving average filtering and outlier removal processing to ensure that the data input to the decision model has a high signal-to-noise ratio and spatiotemporal consistency, thereby providing real-time state boundary conditions for subsequent accurate calculations.

[0066] The pre-processed water level and water temperature data are input into a pre-set collaborative decision-making model. This collaborative decision-making model is constructed based on the ice-water dynamics mechanism and thermodynamic balance equation to simulate the channel response under different control variables. Inside the model, a multi-objective optimization method is introduced in conjunction with the ice damage prevention and control mechanism along the long-distance channel. The ice damage prevention and control mechanism is defined as a thermo-hydraulic coupling criterion to prevent ice flowers from sinking and causing ice blockage.

[0067] During the optimization process, a flow constraint function with the primary objective of suppressing ice cover formation and a water temperature constraint function with the secondary objective of maintaining stable water delivery under the ice cover are established. At the same time, the channel's flow capacity limitation, slope stability conditions, and upstream and downstream hydraulic connection are considered as boundary constraints. By adopting a multi-objective evolutionary approach or a weighted summation method, the flow regulation amplitude and water temperature control threshold are iteratively optimized to find the optimal set of control parameters that meet the current ice control requirements. Thus, under the premise of ensuring the hydraulic safety of the channel, the optimal balance between ice damage suppression effect and water delivery efficiency is achieved.

[0068] Specifically, in response to the "hourly-scale flow pulse strategy" formulated for the high-risk area of ​​the S-shaped bend convex bank, the system immediately triggers the data extraction module. Through the distributed sensor array, it obtains real-time water level data of the S-shaped bend convex bank section, showing that the current water depth is 3.2 meters and the flow velocity is 0.6 m / s. The corresponding vertical water temperature data shows that the surface water temperature has dropped to 0.05 degrees Celsius, close to the critical crystallization point, indicating that the area is on the edge of ice sheet formation.

[0069] The water level and temperature data were input into the collaborative decision-making model. Based on the ice damage prevention mechanism, the model determined that the Froude number (Fr) at the current flow velocity was approximately 0.09, which easily leads to ice crystal accumulation on the convex bank. To suppress ice cover formation, a multi-objective optimization approach, while ensuring channel slope stability, calculated that the Froude number of the convex bank section of the S-curve needed to be instantaneously increased to 0.18 to utilize the shear force of the water flow to break up nascent ice crystals. Based on the optimization results, the system generated specific control parameters:

[0070] Determine the target flow rate change curve: In order to achieve the required flow rate increase at the S-curve, the model inversely determines that the outflow rate of the upstream gate needs to be at the current 30m³. 3 Based on a slope of / s, the slope linearly increases to 36m within 30 minutes according to a preset gradient. 3 / s, forming a flow pulse peak, which falls back to 31m within 60 minutes. 3 / s, this curve accurately describes the pulse formation and decay process.

[0071] Determining the target water temperature control range: In order to coordinate with the flow pulse and prevent the subsequent water from becoming too cold again, the model calculates that the basic heat load of the upstream water needs to be increased, and the target water temperature control range is locked between 0.8 degrees Celsius and 1.2 degrees Celsius. This range can ensure that the water retains enough sensible heat to offset the heat loss along the way during the process of being transported to the S-shaped bend, and can also avoid the water temperature from being too high, which would cause the stable ice cover that has been formed upstream to melt unexpectedly. Thus, under the special hydraulic conditions of the S-shaped channel, the flow disturbance and thermal compensation are precisely coordinated.

[0072] Furthermore, the target flow rate change curve and the target water temperature control range are cross-matched, and the spatiotemporal coupling relationship between hydraulic regulation and thermal compensation is decoupled during the matching process. This spatiotemporal coupling relationship is analyzed, and the coordinated control instructions along the long-distance channel are determined in combination with the current state along the channel. The coordinated control instructions include the timing allocation instructions for upstream reservoir discharge and the operating parameters of each control gate along the channel. The simulation results are introduced to control the multi-scale regulation strategy between different channel sections, thereby improving the accuracy of the coordinated control instructions along the long-distance channel.

[0073] At this point, the target flow rate change curve and the target water temperature control range determined in step S121 are cross-matched in the spatiotemporal dimension to construct a spatiotemporal control matrix containing both flow rate and water temperature attributes. In this process, the complex spatiotemporal coupling relationship between hydraulic control and thermal compensation is identified and decoupled. At this point, the difference between the propagation speed of the flow wave in the channel and the thermal diffusivity of the water body is analyzed to clarify the asynchronous characteristics of the water level fluctuation caused by the flow rate change and the enthalpy change caused by the water temperature change on the time axis. By establishing a time lag difference function, the hydraulic propagation time required for the flow rate regulation command to reach the target channel section and the thermal inertia time required for the water body phase change response caused by the water temperature regulation command are calculated, thereby separating and mapping the hydraulic control target and the thermal compensation target on the time axis.

[0074] Based on this, the spatiotemporal coupling relationship is analyzed, and the independent control variables of the hydraulic-dominated "momentum transport" process and the thermodynamic-dominated "energy transfer" process are extracted. The mutual constraints between the two in the coordinated control are eliminated, ensuring that the flow regulation and water temperature regulation do not interfere with each other and can achieve superposition of effects at specific spatiotemporal nodes.

[0075] Based on the current state along the long-distance channel, the decoupled spatiotemporal coupling relationship is dynamically analyzed, and the optimized control target is transformed into specific control actions. At this time, the current real-time operating conditions of the channel are obtained, including the current opening degree of each gate, the water level difference between upstream and downstream, the operating power of water pumps or heating facilities along the channel, and other state parameters. The ideal target curve determined in step S121 is compared with the current state, and the deviation of the actions to be performed by each control node is calculated.

[0076] For the upstream reservoir discharge stage, the volume of water to be released and its corresponding release time are analyzed to form a time-series allocation scheme. For each control gate stage along the route, the gate opening change rate and specific target opening value required to adapt to the target flow change curve are analyzed during the inversion process. At the same time, a flow and water level dual-factor constraint verification mechanism is introduced to ensure that the analyzed control parameters will not cause the channel water level to rise above the warning level or generate unsteady flow oscillations during execution, thereby generating a coordinated control command with physical feasibility.

[0077] Based on the above analysis results, a complete set of coordinated control instructions covering upstream reservoir discharge and various control gates along the route is generated. This coordinated control instruction is encapsulated into a structured data package containing timing tags, controlled object identifiers, and action parameters. Among them, the timing allocation instruction for upstream reservoir discharge specifies the discharge start time, peak flow duration, and the step division of the flow decay process, aiming to regulate the kinetic energy and total enthalpy of water flow from the source. The operating parameters of each control gate along the route define in detail the target opening degree, opening and closing rate limit, and local flow correction coefficient of each control gate at different time points, aiming to accurately shape the water surface line along the route to adapt to the needs of ice damage prevention. The coordinated control instructions are distributed to the corresponding execution terminals through the industrial control network to achieve a coordinated hydraulic and thermal response of the entire canal system from the water source to the end, ensuring the accurate reproduction of the target flow change curve and the target water temperature control range in the physical channel.

[0078] Specifically, the system cross-matches the target flow rate change curve with the "target water temperature control range, such as 0.8 degrees Celsius to 1.2 degrees Celsius." System analysis reveals that the peak propagation speed of the flow rate pulse is faster than the response speed of water temperature regulation, and the meandering characteristics of the S-shaped channel cause additional lag in the water flow at the bend. Therefore, by decoupling the spatiotemporal coupling relationship between hydraulic regulation and thermal compensation, it is determined that thermal compensation needs to be initiated 30 minutes upstream in advance to increase the water temperature, while hydraulic pulses are initiated 10 minutes later to increase the flow rate, thereby ensuring that the high-temperature water flow and the high-velocity water flow reach the high-risk ice damage zone on the convex bank of the S-shaped bend simultaneously.

[0079] Analysis is conducted based on the current operational status of the S-shaped channel. Currently, the upstream reservoir's spillway gate opening is 2.5 meters, while the openings of control gates #1 and #2 along the channel are 1.8 meters and 1.5 meters, respectively. To achieve the target curve, the system calculates that the upstream reservoir needs to initiate incremental spillway discharge at time T0 and reach a peak flow of 36 m³ / h at T0+30 minutes. 3 / s; Generate specific collaborative control commands:

[0080] The upstream reservoir discharge timing allocation instruction: The instruction requires the upstream reservoir to discharge within the current 30m... 3 Based on a flow rate of / s, the flow rate will be linearly increased starting at 10:00, with the requirement to reach 36m³ / s by 10:30. 3 The peak flow rate was set at 31 m / s, and this flow rate was maintained for 30 minutes before linearly decreasing to 31 m / s starting at 11:00. 3 / s; At the same time, the instruction clearly requires that the temperature of the outflowing water be raised to 1.0 degrees Celsius simultaneously, through deep water intake or mixed heating.

[0081] Operating parameters for each control gate along the line: For control gate No. 1 upstream of the S-shaped curve, the instruction is set that its target opening should be adjusted to 2.2 meters at 10:05, and the opening and closing rate should be limited to no more than 0.1 meters / minute.

[0082] For the No. 2 control gate downstream of the S-shaped bend, the instruction set its target opening degree to be dynamically fine-tuned in conjunction with the passage of the pulse wave. At 10:45, the opening degree was briefly increased to 2.0 meters to increase the discharge capacity and prevent ice blockage on the convex bank from causing the water level to rise. After the pulse passed, the opening degree was restored to 1.6 meters to maintain the subsequent high water level insulation effect. Through the coordinated execution of the above instructions, the precise synchronization of pulse de-icing and thermal insulation in the S-shaped channel was ensured.

[0083] In step S13, the specific steps are as follows:

[0084] S131: Based on the analysis of dynamic coordinated control commands, multiple coordinated control contents are determined, and corresponding linkage regulation is carried out in conjunction with the multi-stage gates along the upstream reservoir and long-distance channels. By controlling the gate opening, flow pulses or water level fluctuations are implemented, and multiple fixed-point disturbance devices are deployed at key bends, inverted siphon inlets, and in front of the gates. The fixed-point disturbance devices adopt a variable frequency disturbance method that combines biomimetic jets and mechanical vibration. The disturbance frequency and amplitude are dynamically adjusted according to the real-time flow velocity vector to break local ice crystals and prevent the initial growth of ice caps.

[0085] S132: The linkage adjustment of heating facilities laid along long-distance channels dynamically adjusts the output power of the heating facilities based on the difference between the target water temperature control range and the measured water temperature, combined with the PID control mechanism, in order to compensate for the sensible and latent heat lost by the water to the atmosphere and ensure that the water temperature is maintained in the dynamic equilibrium range.

[0086] In the embodiments of this application, multiple collaborative control contents are determined based on the parsing of dynamic collaborative control commands, and corresponding linkage adjustments are made in conjunction with multi-stage gates along the upstream reservoir and long-distance channels. By controlling the gate opening, flow pulses or water level fluctuations are implemented, and multiple fixed-point disturbance devices are deployed simultaneously at key bends, inverted siphon inlets, and in front of the gates. The fixed-point disturbance devices adopt a variable frequency disturbance method that combines biomimetic jets and mechanical vibrations. The disturbance frequency and amplitude are dynamically adjusted according to the real-time flow velocity vector to break local ice crystals and prevent the initial growth of ice caps. This approach is compatible with the overall consideration of dynamic collaborative control commands and ensures the accuracy of multiple collaborative control contents.

[0087] At this point, based on the dynamic coordinated control command generated in step S122, the command data packet is unpacked and parsed through the industrial control protocol to extract multiple coordinated control contents, including flow pulse characteristic parameters, water level regulation target values ​​and their time series attributes. Based on the parsed control contents, a cascade response mechanism for multiple gates along the upstream reservoir and long-distance channel is established. According to the propagation characteristics of unsteady flow in open channel, the phase difference and time delay compensation of each gate relative to the regulation action of the upstream reservoir are calculated.

[0088] Through closed-loop feedback control, the hydraulic actuators of each gate are driven to precisely adjust the gate opening, thereby generating and transmitting flow pulses or stabilizing water levels. During this process, the water level difference between upstream and downstream of the gate and the flow rate through the gate are monitored in real time, and the gate opening is finely adjusted to ensure that the waveform distortion rate of the flow pulse wave is controlled within the preset range. Thus, through the coordinated adjustment of all gates, hydraulic boundary conditions that meet the requirements for ice damage prevention are constructed in the channel.

[0089] While the gate is being adjusted in conjunction with the control, multiple fixed-point disturbance devices are simultaneously activated at key bends, inverted siphon inlets, and specific locations in front of the gate, based on the equipment triggering logic determined in the coordinated control command. The deployment locations of the fixed-point disturbance devices are precisely located based on the high-risk ice damage areas identified in step S111, aiming to cover local dead water areas where water flow velocity is slowed down and ice is prone to accumulate.

[0090] At startup, the equipment performs a self-check and enters standby mode, collecting local flow field data near the equipment installation point in real time, including flow velocity, flow direction, and turbulence intensity, to determine the current water flow's carrying capacity for ice crystal transport. By synchronizing the fixed-point disturbance equipment with the gate control system, it ensures that the disturbance effect and the arrival time of the flow pulse peak at the critical node are strictly matched. Thus, on the basis of macro-hydraulic regulation, micro-local dynamic intervention is superimposed to form a full-scale ice damage prevention and control operation system.

[0091] After the fixed-point disturbance equipment enters the operating state, it uses a composite frequency conversion disturbance method that combines biomimetic jet and mechanical vibration to physically destroy local ice crystals. At this time, the biomimetic jet module simulates the vortex flow field generated by the swaying of fish tail fins and emits high-frequency pulse water flow in a specific direction of the water body, generating strong shear stress at the ice-water interface. The mechanical vibration module generates high-frequency micro-amplitude vibration through eccentric motor or piezoelectric ceramics, which directly acts on the ice crystals attached to the surface of the channel wall or structure.

[0092] The adaptive control system inside the equipment dynamically adjusts the disturbance frequency and amplitude based on the real-time flow velocity vector. When a low flow velocity is detected, which makes ice crystals easy to deposit, the jet frequency and vibration intensity are automatically increased to increase the turbulent kinetic energy of the water flow. When a high flow velocity is detected, the disturbance intensity is reduced to save energy. Through this variable frequency disturbance method, the van der Waals forces and hydrogen bonds between ice crystals are destroyed, preventing the initial growth transformation of ice crystals from the flocculated state to the solid ice cap, thereby achieving the elimination and suppression of local ice caps.

[0093] Specifically, the system analyzes the coordinated control instructions and determines that a flow pulse needs to be generated between 10:00 and 11:00. At this time, the dispatch center coordinates and adjusts the No. 1 control gate upstream of the S-shaped channel, increasing its opening from 1.8 meters to 2.2 meters within 10 minutes. At the same time, it coordinates and controls the No. 2 control gate downstream. Through precise gate opening coordination, a pulse waveform of water level first slightly drops and then rises is formed in the S-shaped bend section, using the kinetic energy of water flow to flush away any ice flocs that may be present.

[0094] As the flow pulse wave approaches the high-risk area of ​​the convex bank of the S-shaped bend around 10:45, the system simultaneously activates the fixed-point disturbance equipment deployed at the bottom of the convex bank slope and in front of the inverted siphon inlet. The fixed-point disturbance equipment activates the "bionic jet + mechanical vibration" mode. The sensors on the equipment detect that the flow velocity vector in the convex bank backflow area shows that the flow velocity is only 0.15m / s and the direction is disordered, which is very likely to cause ice crystal deposition. Based on this, the equipment adaptively adjusts the disturbance parameters: the bionic jet device emits high-pressure water flow downwards at a frequency of 25Hz, simulating the directional vortex generated by biological propulsion, and interrupting the growing ice crystal network.

[0095] Meanwhile, the mechanical vibration device excites the surrounding water body at a frequency of 50Hz and an amplitude of 2mm. As the flow pulse peak arrives, the flow velocity gradually increases to 0.8m / s. The equipment automatically reduces the disturbance frequency to 10Hz and switches to the auxiliary flow mode. Through this frequency conversion adjustment, the structure of the nascent ice crystals on the convex bank of the S-shaped bend is effectively destroyed, preventing the ice cover from growing laterally along the inner side of the bend and ensuring ice-free water transport in the S-shaped channel during this period.

[0096] Furthermore, the heating facilities laid along the long-distance channel are linked and adjusted. Based on the difference between the target water temperature control range and the measured water temperature, and combined with the PID control mechanism, the output power of the heating facilities is dynamically adjusted to compensate for the sensible and latent heat lost by the water to the atmosphere, so as to ensure that the water temperature is maintained in the dynamic equilibrium range.

[0097] At this time, based on the target water temperature control range determined in step S121 and the measured water temperature data collected by distributed temperature sensors along the long-distance channel during the execution of step S131, the real-time deviation between the two is calculated as the input variable for thermal compensation; when the deviation is less than zero or lower than the preset lower limit threshold, the linkage adjustment mechanism of the heating facilities laid along the long-distance channel is triggered.

[0098] During this process, the system identifies the heating unit zone corresponding to the channel section currently in a state of thermal imbalance based on the ice distribution map and hydraulic flow direction, and sends a linkage start signal to the thyristor trigger or power regulation module of that zone; the heating facilities respond to the signal and gradually start operation according to the preset soft start strategy to avoid instantaneous large current impact on the power grid, while rapidly transferring heat energy to the water body through the heat conduction medium to compensate for the sensible heat and latent heat of vaporization lost by the water body to the atmosphere due to the low temperature environment, thereby curbing the continuous downward trend of water temperature.

[0099] After the heating facility is put into operation, the system introduces a proportional-integral-derivative (PID) control mechanism to dynamically adjust the output power of the heating facility in a closed loop. At this time, the corresponding power correction factor is calculated based on the magnitude of the real-time temperature difference, the historical trend of the accumulated temperature difference, and the rate of change of water temperature. The proportional stage adjusts the base heating power proportionally according to the current temperature difference to ensure a rapid response to large temperature drops. The integral stage integrates the historical temperature difference over time to eliminate steady-state errors and prevent the water temperature from remaining below the lower limit of the target range for a long time. The derivative stage makes predictive adjustments based on the slope of the water temperature change to suppress overshoot oscillations in the water temperature during the heating process. Through the coordinated operation of the above three stages, the system generates precise power control commands to drive the actuators of the heating facility to make real-time adjustments, achieving a precise match between heating output and heat load demand.

[0100] Based on the dynamic adjustment power output of the PID control mechanism, the heating facilities continuously inject heat energy into the water body of the canal section until the measured water temperature re-enters and stabilizes within the target water temperature control range. During this process, the system continuously monitors the changes in heat exchange coefficient caused by "environmental meteorological parameters, such as sudden changes in temperature or increased wind speed," and corrects the set parameters of the PID controller in real time to adapt to external thermal disturbances. By continuously balancing the heat lost from the water body with the heat replenished by the heating facilities, the system maintains the water temperature within a dynamic equilibrium range, ensuring that the water temperature is slightly above the freezing point without overheating, thereby effectively preventing the water body from freezing while minimizing the overall operating energy consumption of the heating facilities. At the same time, the system records the operating power and temperature rise effect of each heating unit, establishes an energy efficiency evaluation model, and uses it to optimize the self-tuning of PID parameters in subsequent control cycles, ensuring the long-term stable operation of the long-distance canal thermal compensation system.

[0101] Optionally, the heating system uses an armored electric heating cable network continuously laid along the bottom of the canal as the heating element. The rated power of the heating cable network is 200 watts per meter. It is divided into an independent control zone every 50 meters. Each zone is equipped with a set of thyristor power adjustment modules and an armored temperature sensor with an accuracy of ±0.1 degrees Celsius. The temperature sensor is installed in the water 20 centimeters above the bottom of the canal. The entire heating control system uses a programmable logic controller as the central processing unit. The controller reads the measured water temperature value of the temperature sensor in each zone at a sampling frequency of once per second.

[0102] The implementation process of the PID control mechanism is as follows: The system converts the target water temperature control range into a single target setpoint temperature value. Specifically, the midpoint of the target water temperature control range is taken as the setpoint. For example, when the target water temperature control range is 0.8 degrees Celsius to 1.2 degrees Celsius, the target setpoint temperature is 1 degree Celsius. Then, the controller calculates the real-time deviation between the measured water temperature and the target setpoint temperature in each control cycle. This deviation is equal to the measured water temperature minus the target setpoint temperature, in degrees Celsius. Simultaneously, the controller also records the rate of change of the deviation, i.e., the change in deviation between two adjacent control cycles divided by the control cycle time.

[0103] The percentage of heating power output by the PID controller is calculated using the following formula: Output power percentage equals proportional gain multiplied by the current deviation value, plus integral gain multiplied by the cumulative integral of the deviation value over time, plus derivative gain multiplied by the rate of change of the deviation value. The proportional gain is set to 120, meaning that when the deviation is -0.5 degrees Celsius, the proportional term contributes -60% of the output adjustment. The integral gain is set to 0.5 per second; the integral term is used to eliminate static errors, and its output value equals the integral gain multiplied by the cumulative sum of deviation values ​​over the past five minutes, divided by 300 seconds. The derivative gain is set to 40 seconds; the derivative term is used to predict the water temperature change trend, and its output value equals the derivative gain multiplied by the difference between the current deviation value and the deviation value of the previous cycle, divided by the control cycle. After summing these three terms, the output is limited to between 5% and 100% through output limiting processing. 5% is the minimum power output for maintaining standby mode, used to prevent backflow after slight over-adjustment of the water temperature.

[0104] Specifically, for the high-risk area of ​​ice damage on the convex bank of the S-shaped bend, the system sets the target water temperature control range to 0.8 degrees Celsius to 1.2 degrees Celsius. While executing flow pulse disturbance, the system obtains the measured water temperature through distributed fiber optic sensors laid at the bottom of the S-shaped bend, which has dropped to 0.45 degrees Celsius, and is 0.35 degrees Celsius lower than the lower limit of the target range. Based on this temperature difference, the system determines that the sensible heat loss of the water to the cold atmosphere is too fast, and immediately triggers the linkage adjustment of the electric heating cable network located at the bottom of the channel section.

[0105] The PID control mechanism is activated for power adjustment. Due to the large current temperature difference (e.g., 0.35 degrees Celsius), the proportional element in the PID controller outputs a larger base power, requiring the electric heating cable network to operate at 80% of its rated power to quickly curb the water temperature drop. Simultaneously, the integral element calculates the continuous heat deficit over the past 10 minutes and adds an increment to the control output to ensure the water temperature recovers to the set value. The derivative element detects the rate of water temperature decrease. The system is set at 0.05 degrees Celsius / min. It is predicted that the water temperature will continue to drop without additional intervention, so the power output is further fine-tuned to increase. The system converts the calculated total power command into a voltage control signal and dynamically adjusts the current input of the electric heating cable network so that its output power fluctuates with the real-time temperature difference.

[0106] As the heating system continues to operate, the water temperature on the convex bank of the S-bend begins to rise. When the water temperature rises to 0.9 degrees Celsius and enters the target range, the PID controller automatically reduces the output power and switches to maintenance mode, for example, reducing the power to 30% to compensate for the natural heat dissipation of the water in the low-temperature environment. After a period of dynamic adjustment, the water temperature stabilizes at around 1.0 degrees Celsius, ensuring that the water body, in conjunction with the flow pulse, has sufficient sensible heat to melt the tiny ice crystals, while avoiding overheating and energy waste. This successfully achieves the thermal balance and anti-icing target of the local water body in the S-bend.

[0107] In step S14, the specific steps are as follows:

[0108] S141: Camera terminals deployed along long-distance channels collect ice condition images along the channels and simultaneously acquire corresponding hydraulic element data. The ice condition images are then image-recognized, and combined with a preset deviation assessment mechanism, the actual ice condition evolution trend and the projection results are subjected to multi-dimensional residual analysis and consistency verification to determine the deviation between the actual ice condition and the projection results. The deviation includes the time lag deviation, positional offset deviation, and morphological structure deviation of the ice cover formation.

[0109] S142: Tracing back the deviation content and establishing the attribution mapping relationship between the deviation elements and different meteorological and hydraulic boundary conditions during the tracing process, thereby identifying multiple ice damage prevention projects. These projects cover upstream preheating correction, local disturbance enhancement, and gate control optimization. The impact weights along the long-distance channel are ranked based on the urgency of each ice damage prevention project, thereby determining the ice damage prevention sequence of the water conveyance channel.

[0110] In the embodiments of this application, camera terminals deployed along long-distance channels collect ice condition images along the channels and simultaneously acquire corresponding hydraulic element data. The ice condition images are then image-recognized, and combined with a preset deviation evaluation mechanism, the actual ice condition evolution trend and the projected results are subjected to multi-dimensional residual analysis and consistency verification. This determines the deviation between the actual ice condition and the projected results. The deviation includes the time lag deviation, positional offset deviation, and morphological structure deviation of ice cover formation. The time lag deviation, positional offset deviation, and morphological structure deviation of ice cover formation are introduced.

[0111] At this time, high-definition camera terminals and infrared thermal imaging equipment deployed along the long-distance canal scan the canal surface around the clock according to the preset sampling frequency, collecting ice condition images including ice flower distribution, ice cover edges, and surface texture features. Simultaneously, the system obtains hydraulic element data at the corresponding moment from hydrological sensors deployed at the same cross-section through an industrial bus and IoT synchronization interface. This hydraulic element data includes key parameters such as water level, flow velocity, flow rate, and water temperature. To ensure the accuracy of subsequent analysis, the system performs timestamp calibration and spatial coordinate mapping on the collected image data, strictly aligning and registering visual and hydraulic information in time and space, constructing a multimodal dataset containing visual and physical features, which serves as the basic data source for assessing the actual ice condition evolution trend.

[0112] A deep learning-based computer vision approach is used to perform semantic segmentation and target detection on ice condition images, thereby extracting the geometric and texture features of the actual ice conditions. In this process, a trained convolutional neural network model is used to identify ice, water, and shore ice boundaries in the images, and to calculate the current ice cover coverage, ice edge position, and ice surface roughness.

[0113] Meanwhile, by combining infrared thermal imaging data to invert the temperature field distribution on the ice sheet surface, and by performing differential processing and optical flow analysis on continuous time series image frames, the advancing speed of the ice sheet front, the movement trajectory of the ice flow, and the dynamic change trend of the ice sheet thickness are captured, thereby generating a spatiotemporal state vector that characterizes the actual ice condition evolution trend. This vector quantitatively describes the spatial distribution of the ice condition at the current moment and its temporal development rate.

[0114] Combined with the preset deviation assessment mechanism, the actual ice condition evolution trend is compared and analyzed with the inference results output in step S111 in multiple dimensions. At this time, the spatiotemporal state vector of the actual ice condition and the predicted state vector output by the inference model are input into the residual calculation module to calculate the residual values ​​of ice condition formation time, spatial location and morphological characteristics respectively. By setting the confidence interval and consistency test threshold, it is determined whether the inference results match the actual observation values.

[0115] Optionally, the deviation assessment mechanism is implemented through multidimensional residual analysis, specifically including the following three calculation methods: The residual calculation method for the time dimension is as follows: the time lag deviation equals the actual ice sheet formation time minus the predicted ice sheet formation time, in minutes. A positive value indicates that the actual formation is later than the prediction, and a negative value indicates that the actual formation is earlier than the prediction. The system sets a time consistency test threshold of ±15 minutes; that is, when the absolute value of the time lag deviation exceeds 15 minutes, it is judged as a significant deviation in the time dimension.

[0116] The residual calculation method in the spatial dimension is as follows: the position offset deviation is equal to the actual ice edge coordinates minus the predicted ice edge coordinates, and the unit is meters; the spatial consistency test threshold set by the system is ±20 meters, that is, when the absolute value of the position offset deviation exceeds 20 meters, it is judged as a significant deviation in the spatial dimension.

[0117] The residual calculation method for the morphological dimension is as follows: First, the difference between the actual ice sheet thickness and the predicted ice sheet thickness at each measuring point is calculated to obtain the thickness residual distribution curve. Then, the root mean square error of this curve over all measuring points is calculated and defined as the morphological deviation, in centimeters. The system sets the morphological consistency test threshold to 3 centimeters, meaning that when the morphological deviation exceeds 3 centimeters, it is judged as a significant deviation in the morphological dimension. When the absolute value of the residual in any of the above three dimensions exceeds the corresponding test threshold, the system triggers the deviation identification process and outputs a deviation report.

[0118] If the residual value exceeds the preset threshold range, it is determined to be inconsistent and the deviation identification process is triggered; if the residual value is within the threshold range, it is determined to be consistent and the system maintains the current inference model parameters. This process aims to quantitatively assess the uncertainty of the model prediction and identify the cumulative error in the inference process or the model distortion caused by sudden changes in the external environment.

[0119] Based on the results of multidimensional residual analysis, the deviations between the actual ice conditions and the projected results are quantitatively classified and determined. In the time dimension, the time difference between the actual ice cover formation time and the projected prediction time is compared to determine the time lag deviation and quantify the model's response speed. In the spatial dimension, the Euclidean distance between the actual ice edge line or ice block center point coordinates and the projected coordinates is calculated. Combined with the channel mileage markers, the positional offset deviation is determined, and the spatial error distribution of the model is identified. In the morphological dimension, the thickness, surface smoothness, and ice flower accumulation structure characteristics of the actual ice cover are extracted and compared with the projected morphological structure to determine the morphological structure deviation. This identifies the accuracy of the model's description of ice flower flocculation, accumulation, and the mechanical stability of the ice cover. Through the determination of deviations in the above three dimensions, a complete deviation analysis report is generated, providing accurate feedback for subsequent model correction and control strategy optimization.

[0120] Specifically, the high-definition camera terminal deployed on the convex bank of the S-shaped bend acquired a set of high-resolution channel surface images at 11:00. At the same time, it simultaneously acquired hydraulic element data of the convex bank section at that moment: water level of 3.25 meters, flow velocity of 0.65 m / s, and water temperature of 1.05 degrees Celsius. The system performs millisecond-level time alignment between the image data and the hydraulic data to construct an ice condition data package for the S-shaped bend at that moment.

[0121] The data packet was processed using image recognition. A large number of grayish-white flocculent materials were identified on the inner side of the S-shaped bend in the image, which were determined to be high-density ice flower accumulation. The ice flower accumulation zone showed an irregular strip-shaped distribution along the direction of water flow. Through continuous frame analysis, the system calculated that the actual ice condition showed that the ice flowers were rapidly converging towards the center of the cross section, and thin ice had begun to form on the bank. The actual ice condition evolution trend pointed to "a local ice blockage is about to occur".

[0122] The actual trend described above was compared with the projection results in step S111 to assess the discrepancy. The projection results originally predicted that a stable ice sheet would form at 11:30, with its location mainly concentrated at a distance of 22 meters from the shoreline. However, through multidimensional residual analysis, it was found that signs of ice sheet formation actually appeared at 11:00, 30 minutes earlier than the projection results. The time lag bias was determined to be [missing value]. 30 minutes; the actual ice flower accumulation core was located 3.5 meters from the shoreline, biased towards the center of the channel, 1.5 meters off from the projected location, with the determined position deviation being +1.5 meters; in addition, the actual ice flower structure exhibited a loose network rather than the dense layered structure in the projected model, with the determined morphological structure deviation being "high structural looseness"; based on these determined deviations, the system determined that the projected model underestimated the transport effect of the S-shaped bend circulation on the ice flowers, thus providing a direction for correction in subsequent strategy adjustments.

[0123] Furthermore, the deviations were traced back, and during the tracing process, attribution mapping relationships between deviation elements and different meteorological and hydraulic boundary conditions were established. This led to the identification of multiple ice damage prevention projects, which covered upstream preheating correction, local disturbance enhancement, and gate control optimization. The impact weights along long-distance channels were ranked based on the urgency of each ice damage prevention project, thereby determining the ice damage prevention sequence for the water conveyance channel. The fixed-point disturbance equipment and heating facilities were further controlled, and the deviations between the actual ice condition evolution trend and the projection results were fully considered, thus improving the accuracy of the ice damage prevention sequence for the water conveyance channel.

[0124] At this point, based on the deviation content determined in step S141, a deep feature extraction is performed to transform the qualitative deviation description into a quantifiable deviation factor. During this process, the system traverses the time window and spatial location of the deviation occurrence, and simultaneously retrieves the "meteorological conditions, such as sudden drop in temperature, sudden change in wind speed and direction, and change in radiative heat" and "hydraulic boundary conditions, such as upstream flow fluctuation, gate opening amplitude, and water level variability" corresponding to the spatiotemporal node. By constructing a multidimensional correlation matrix, the extracted deviation factors are spatiotemporally aligned and matched with specific environmental parameters to establish a preliminary mapping relationship between deviation elements and environmental parameters, aiming to identify the external triggering sources and internal state variables that cause the deviation.

[0125] Based on the initial mapping, the causes of deviations are logically reasoned and inverted using the ice-hydraulic mechanism model, establishing the attribution mapping relationship between deviation elements and different meteorological and hydraulic boundary conditions. At this point, it is analyzed whether the deviation is caused by meteorological and thermodynamic conditions exceeding the model's prediction range or by changes in hydraulic boundary conditions. Based on the attribution results, corresponding governance measures are retrieved from the pre-set prevention and control strategy library, thereby identifying multiple targeted ice damage prevention and control projects. This project set includes: "upstream preheating correction" for insufficient heat input, "local disturbance enhancement" for local flow field ice dynamic instability, and "gate control optimization" for insufficient hydraulic ice transport capacity. Each project addresses ice damage risks under a specific attribution.

[0126] For the identified ice damage prevention projects, their impact on the operational safety of long-distance channels and the urgency of their implementation are quantitatively assessed. A multi-objective evaluation system is established, comprehensively considering the ice damage risk level, deviation diffusion rate, and impact weight on water conveyance efficiency of each project's coverage area, and calculating the comprehensive urgency index for each project. Based on the urgency index, all ice damage prevention projects are ranked in descending order, and a temporal and spatial mutual exclusion and synergistic constraint logic is introduced between projects to eliminate projects that conflict in time or space, thereby determining the final ice damage prevention sequence for the water conveyance channel. This ice damage prevention sequence specifies the order of implementation, start and end times, and resource allocation priority of different prevention projects, ensuring that control measures can be implemented in an orderly manner according to the risk level, maximizing the benefits of ice damage prevention.

[0127] Optionally, for multiple identified snow and ice prevention projects, the system calculates a comprehensive urgency index for each project. This comprehensive urgency index is obtained by weighted summation of three sub-indicators. The first sub-indicator is the risk level coefficient, assigned a value of 0.8 for high-risk snow and ice areas, 0.4 for transitional stable areas, and 0.1 for ice-free water conveyance areas. The second sub-indicator is the deviation diffusion rate coefficient, calculated by dividing the increase in hourly time lag deviation by 60 minutes to obtain the deviation diffusion rate per minute. A value of 0.7 is assigned when the diffusion rate is greater than 0.5 minutes, 0.3 when it is between 0.1 and 0.5 minutes per minute, and 0.1 when it is less than 0.1 minutes per minute. The third sub-indicator is the water conveyance impact coefficient, assigned a value of 0.9 when the covered channel section is a water conveyance bottleneck structure such as an inverted siphon or aqueduct, and 0.5 when it is an ordinary open channel section. The comprehensive urgency index is calculated using the following formula: the comprehensive urgency index equals 0.4 multiplied by the risk level coefficient plus 0.3 multiplied by the deviation diffusion rate coefficient plus 0.3 multiplied by the water conveyance impact coefficient. The sum of the three weights is one. The calculated comprehensive urgency index ranges from 0.1 to 0.85. The higher the value, the higher the urgency level.

[0128] After calculating the overall urgency of each project, the system sorts all snow and ice prevention projects in descending order of their numerical values ​​to obtain an initial sorting sequence. The system then performs spatiotemporal conflict resolution on this initial sorting sequence. The conflict resolution rules are as follows: when the execution areas of two prevention projects overlap spatially and their execution time windows overlap, the project with the higher overall urgency is retained, and the project with the lower overall urgency is postponed. The minimum postponement time interval is one execution cycle, i.e., thirty minutes.

[0129] When the execution areas of two snow and ice prevention projects are not spatially adjacent, parallel execution is allowed, and the start times of both are aligned to a unified clock reference. After conflict resolution, the system outputs the final snow and ice prevention sequence, which is presented in list form. Each row in the list contains three pieces of information: project name, execution start time, and execution priority number. The priority number is 1, which indicates the highest priority, and increases sequentially. The system triggers the corresponding prevention and control projects in the sequence according to this sequence. After each project is completed, the next project in the sequence is automatically triggered until all projects are completed, thereby achieving the orderly and efficient deployment of snow and ice prevention measures.

[0130] Specifically, the system extracts the elements identified in step S141, namely "S-shaped bend deviation: 30-minute time lag, 1.5-meter positional offset, and loose structure," and links them with environmental data from 10:30 to 11:00. The analysis reveals that during this period, the local wind speed at the S-shaped bend suddenly increased to 88 m / s, and the upstream No. 1 gate had previously performed a rapid flow pulse adjustment, such as hydraulic boundary conditions, which resulted in the transverse circulation intensity of the water surface at the bend being significantly higher than the model prediction.

[0131] Based on the above analysis, an attribution mapping relationship was established: the system determined that "positional shift" and "loose structure" were due to strong winds accelerating the transport of surface ice to the shore, and the unexpected circulation intensity engulfing ice flakes into the core of the backflow zone, which belong to the hydrodynamic boundary attribution; while "time lag" was due to strong winds carrying away more sensible heat, resulting in a slower cooling rate predicted by the thermal model, which belongs to the meteorological and thermal condition attribution; based on this, the system identified three ice damage prevention projects: "upstream preheating correction, such as increasing heat input" for insufficient upstream water temperature, "local disturbance enhancement, increasing the frequency of disturbance equipment on the convex bank" for S-shaped bend circulation accumulation, and "gate control optimization, adjusting gate #2 to change the downstream water level gradient of the bend" for limited ice transport capacity of the bend.

[0132] The system prioritizes tasks based on urgency. Since the accumulation of ice crystals on the convex bank of the S-curve directly threatens the channel's flow section and could easily cause ice blockage and flow interruption, "local disturbance enhancement" is determined to be the highest priority. Meanwhile, to prevent the formation of new ice, "upstream preheating correction" is listed as the second priority. "Gate control optimization," due to its involvement in water level fluctuations across the entire channel, requires careful operation and is therefore listed as the third priority. Based on this, the system generates the final ice damage prevention sequence: immediately increase the power of the disturbance equipment on the convex bank of the S-curve (first priority); 10 minutes later, initiate the deep water intake preheating program of the upstream reservoir (second priority); 15 minutes later, fine-tune the opening of gate #2 (third priority), thereby systematically mitigating the ice damage risk of the S-shaped channel.

[0133] Please see Figure 2 The winter long-distance water conveyance channel ice damage prevention system is applied to the above-mentioned winter long-distance water conveyance channel ice damage prevention method; the winter long-distance water conveyance channel ice damage prevention system includes:

[0134] The water conveyance channel module 21 is used to acquire multimodal data along the long-distance channel during the winter of the water conveyance channel, to extrapolate the spatiotemporal evolution of the heat exchange process and ice cover formation process along the long-distance channel, and to output the extrapolation results. Based on the identification of the extrapolation results, multi-scale control strategies between different channel sections are determined.

[0135] The collaborative control module 22 is used to respond to the multi-scale regulation strategy, and simultaneously extract the water level data and water temperature data along the long-distance channel at the current moment, and combine the ice damage prevention and control mechanism along the long-distance channel to determine the target flow change curve and the target water temperature control range, thereby generating collaborative control instructions along the long-distance channel.

[0136] The linkage adjustment module 23 is used to trigger the linkage adjustment of the gates along the upstream and long-distance channels based on the dynamic collaborative control command, and simultaneously start multiple fixed-point disturbance devices to break local ice crystals in the form of frequency conversion disturbance, while adjusting the output power of the heating facilities to compensate for the heat loss of the water.

[0137] The ice damage prevention sequence module 24 is used to collect ice condition images and hydraulic element data along long-distance channels, and combine them with a preset deviation assessment mechanism to determine the deviation between the actual ice condition evolution trend and the projection results. Based on the traceability of the deviation, multiple ice damage prevention projects are determined, thereby determining the ice damage prevention sequence of the water conveyance channel.

[0138] It should be noted that although multiple modules are mentioned in the detailed description above, this division is not mandatory; in fact, according to the embodiments of this disclosure, the features and functions of two or more modules or described above can be embodied in one module; conversely, the features and functions of one module described above can be further divided into multiple modules to be embodied.

[0139] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure herein; this application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein; the specification and embodiments are to be considered exemplary only.

[0140] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for preventing ice damage in long-distance water conveyance channels during winter, characterized in that, include: During the winter season, multimodal data along the long-distance water conveyance channel is acquired to extrapolate the spatiotemporal evolution of the heat exchange process and ice cover formation process along the long-distance channel, thereby outputting the extrapolation results. Based on the identification of the extrapolation results, multi-scale control strategies between different channel sections are determined. In response to the multi-scale control strategy, the system simultaneously extracts water level and temperature data along the long-distance channel at the current moment, and combines the ice damage prevention and control mechanism along the long-distance channel to determine the target flow change curve and the target water temperature control range, thereby generating collaborative control instructions along the long-distance channel. Based on the dynamic collaborative control command, the gates along the upstream and long-distance channels are linked and adjusted, and multiple fixed-point disturbance devices are started simultaneously to break local ice crystals by frequency conversion disturbance. At the same time, the output power of the heating facilities is adjusted to compensate for the heat loss of the water. Ice condition images and hydraulic element data are collected along long-distance channels. Combined with a pre-set deviation assessment mechanism, the deviation between the actual ice condition evolution trend and the projection results is determined. Based on the traceability of the deviation, multiple ice damage prevention and control projects are determined, thereby determining the ice damage prevention and control sequence for the water conveyance channel.

2. The method for preventing ice damage in long-distance water conveyance channels during winter according to claim 1, characterized in that, During the winter season of the water conveyance channel, multimodal data along a long distance of the channel is acquired to extrapolate the spatiotemporal evolution of the heat exchange process and ice cover formation process along the long distance of the channel, thereby outputting the extrapolation results. Based on the identification of these extrapolation results, multi-scale control strategies for different channel sections are determined, including: For the winter season of water conveyance channels, multimodal data are collected along long-distance channels. The multimodal data includes meteorological field data, water temperature distribution data, flow velocity field data, and ice cover thickness point cloud data. Dynamic simulations are performed on the multimodal data, and the heat exchange process along the long-distance channels is inverted in real time. In addition, the spatiotemporal evolution of the ice cover formation process is simulated by combining ice-hydraulic dynamics, and the simulation results are output, which include the spatiotemporal distribution characteristics of ice condition evolution.

3. The method for preventing ice damage in long-distance water conveyance channels during winter according to claim 2, characterized in that, During the winter season of the water conveyance channel, multimodal data along a long distance of the channel is acquired to extrapolate the spatiotemporal evolution of the heat exchange process and ice cover formation process along the long distance of the channel, thereby outputting the extrapolation results. Based on the identification of these extrapolation results, multi-scale control strategies between different channel sections are determined. The method also includes: Based on the identification of the simulation results, the corresponding ice flower density and ice flow volume are determined. Combined with the distribution images along the long-distance channel, the corresponding high-risk area for ice damage, transitional stable area and ice-free water conveyance area are determined. Based on the thermal imbalance state of the high-risk area for ice damage, transitional stable area and ice-free water conveyance area, multi-scale control strategies between different channel sections are determined. The multi-scale control strategies include flow pulse strategy on an hourly scale and heat storage and insulation strategy on a daily scale.

4. The method for preventing ice damage in long-distance water conveyance channels during winter according to claim 1, characterized in that, The response to the multi-scale control strategy simultaneously extracts water level and temperature data along the long-distance channel at the current moment, and combines this with the ice damage prevention mechanism along the long-distance channel to determine the target flow rate change curve and the target water temperature control range, thereby generating coordinated control instructions along the long-distance channel, including: In response to the multi-scale control strategy, the system extracts the current water level and water temperature data from the distributed sensor array along the long-distance channel. The water level and water temperature data are then input into the preset collaborative decision-making model. Combined with the ice damage prevention and control mechanism along the long-distance channel and the corresponding constraint relationship, multi-objective optimization is performed to determine the target flow change curve for suppressing ice cover formation and the target water temperature control range for maintaining stable water delivery under the ice cover.

5. The method for preventing ice damage in long-distance water conveyance channels during winter according to claim 4, characterized in that, The response to the multi-scale control strategy, which simultaneously extracts water level and temperature data along the long-distance channel at the current moment, and combines this with the ice damage prevention mechanism along the long-distance channel to determine the target flow rate change curve and the target water temperature control range, thereby generating coordinated control commands along the long-distance channel, also includes: Cross-matching is performed between the target flow rate change curve and the target water temperature control range. During the matching process, the spatiotemporal coupling relationship between hydraulic regulation and thermal compensation is decoupled. This spatiotemporal coupling relationship is analyzed. During the analysis process, the coordinated control instructions along the long-distance channel are determined in conjunction with the current state along the channel. The coordinated control instructions include the timing allocation instructions for upstream reservoir discharge and the operating parameters of each control gate along the channel.

6. The method for preventing ice damage in long-distance water conveyance channels during winter according to claim 1, characterized in that, The linkage adjustment of gates along the upstream and long-distance channels triggered by dynamic collaborative control commands simultaneously activates multiple fixed-point disturbance devices to break local ice crystals through frequency conversion disturbance, while adjusting the output power of heating facilities to compensate for heat loss from the water, including: Based on the analysis of dynamic coordinated control commands, multiple coordinated control contents are determined, and corresponding linkage regulation is carried out in conjunction with the multi-stage gates along the upstream reservoir and long-distance channels. By controlling the gate opening, flow pulses or water level fluctuations are implemented, and multiple fixed-point disturbance devices are deployed at key bends, inverted siphon inlets, and upstream of the gates. The fixed-point disturbance devices adopt a variable frequency disturbance method that combines biomimetic jets and mechanical vibration. The disturbance frequency and amplitude are dynamically adjusted according to the real-time flow velocity vector to break local ice crystals and prevent the initial growth of ice caps.

7. The method for preventing ice damage in long-distance water conveyance channels during winter according to claim 6, characterized in that, The method, based on dynamic collaborative control commands triggering the coordinated adjustment of gates along the upstream and long-distance channels, simultaneously activates multiple fixed-point disturbance devices to break local ice crystals through frequency conversion disturbance, and adjusts the output power of heating facilities to compensate for heat loss from the water, also includes: The heating facilities laid along long-distance channels are linked and adjusted. Based on the difference between the target water temperature control range and the actual water temperature, and combined with the PID control mechanism, the output power of the heating facilities is dynamically adjusted to compensate for the sensible and latent heat lost by the water to the atmosphere, so as to ensure that the water temperature is maintained in the dynamic equilibrium range.

8. The method for preventing ice damage in long-distance water conveyance channels during winter according to claim 1, characterized in that, The process involves collecting ice condition images and hydraulic element data along long-distance channels, combining this data with a pre-set deviation assessment mechanism to determine the discrepancies between the actual ice condition evolution trend and the projected results, and identifying multiple ice damage prevention projects based on the traceability of these discrepancies. This determines the ice damage prevention sequence for the water conveyance channel, including: Camera terminals deployed along long-distance channels collect ice condition images along the channels and simultaneously acquire corresponding hydraulic element data. The ice condition images are then used for image recognition. Combined with a pre-set deviation assessment mechanism, the actual ice condition evolution trend and the projected results are subjected to multi-dimensional residual analysis and consistency verification. This determines the deviation between the actual ice condition and the projected results. The deviation includes the time lag deviation of ice cover formation, the positional offset deviation, and the morphological structure deviation.

9. The method for preventing ice damage in long-distance water conveyance channels during winter according to claim 8, characterized in that, The process of collecting ice condition images and hydraulic element data along long-distance channels, combining this with a pre-set deviation assessment mechanism to determine the deviation between the actual ice condition evolution trend and the projected results, and identifying multiple ice damage prevention projects based on the traceability of these deviations, thereby determining the ice damage prevention sequence for the water conveyance channel, also includes: The deviations were traced back, and during the tracing process, the attribution mapping relationship between the deviation elements and different meteorological and hydraulic boundary conditions was established, thereby identifying multiple ice damage prevention projects. These projects cover upstream preheating correction, local disturbance enhancement, and gate control optimization. Based on the urgency of each ice damage prevention project, the impact weights along the long-distance channel were ranked, thus determining the ice damage prevention sequence of the water conveyance channel.

10. A system for preventing ice damage in long-distance water conveyance channels during winter, characterized in that, The winter long-distance water conveyance channel ice damage prevention system is applied to the winter long-distance water conveyance channel ice damage prevention method as described in any one of claims 1-9; the winter long-distance water conveyance channel ice damage prevention system includes: The water conveyance channel module is used to acquire multimodal data along the long-distance water conveyance channel during winter, to extrapolate the spatiotemporal evolution of the heat exchange process and ice cover formation process along the long-distance channel, and to output the extrapolation results. Based on the identification of the extrapolation results, multi-scale control strategies between different channel sections are determined. The collaborative control module is used to respond to multi-scale regulation strategies, while extracting water level and water temperature data along the long-distance channel at the current moment, and combining the ice damage prevention and control mechanism along the long-distance channel to determine the target flow change curve and the target water temperature control range, thereby generating collaborative control instructions along the long-distance channel. The linkage adjustment module is used to trigger the linkage adjustment of gates along the upstream and long-distance channels based on dynamic collaborative control commands, and simultaneously start multiple fixed-point disturbance devices to break local ice crystals by frequency conversion disturbance, while adjusting the output power of heating facilities to compensate for the heat loss of the water. The ice damage prevention sequence module is used to collect ice condition images and hydraulic element data along long-distance channels. Combined with a preset deviation assessment mechanism, it determines the deviation between the actual ice condition evolution trend and the projection results. Based on the traceability of the deviation, multiple ice damage prevention projects are determined, thereby determining the ice damage prevention sequence of the water conveyance channel.

Images