Intelligent data monitoring analysis method and system of solar seawater desalination system

Abstract

The invention provides an intelligent data monitoring analysis method and system of a solar seawater desalination system. The method comprises the following steps: calculating condensation efficiency based on annular temperature gradient distribution data and phase change characteristic data according to the phase change characteristic data and environment temperature data of steam in a condensation pipe in a condensation process; the condensation efficiency and the environment temperature data are input into a pre-trained reinforcement learning parameter optimization model, and in combination with physical constraint conditions of the condensation structure, flow adjusting parameters and deflector deflection angle adjusting parameters of all subareas are generated; and the flow adjusting parameters and the deflector deflection angle adjusting parameters of all the partitions are converted into corresponding electric adjusting valve opening degree instructions, and the adjusting valve opening degrees of the cooling water circulation systems corresponding to the different partitions and the angle of the angle-adjustable deflector are adjusted. The cooling water flow and the deflection angle of the guide plate are optimized in real time through reinforcement learning, and efficient and energy-saving self-adaptive regulation and control of the condensation system are achieved through temperature and phase change data monitoring.

Description

Technical Field

[0001] This application relates to the technical field of seawater desalination, and particularly to an intelligent data monitoring and analysis method and system for a solar seawater desalination system. Background Art

[0002] The condensation stage of a solar seawater desalination system is a key link determining the overall energy efficiency. Its technical requirements focus on dynamically monitoring and intelligently regulating the thermodynamic state of a multi-zone condensation structure. Due to the strong coupling relationship among the annular temperature gradient distribution of the condensation tube wall, the steam phase change characteristics, and the environmental temperature change, it is necessary to collect multi-dimensional data in real time for this system and dynamically adjust the cooling water flow rate and the deflection angle of the baffle plate based on this to balance the condensation efficiency and energy consumption. Especially under complex working conditions such as fluctuating light intensity and changing seawater salinity, traditional static control strategies are difficult to adapt to dynamic thermal loads, and there is an urgent need for a method that can autonomously sense, model, and optimize the condensation process to achieve adaptive operation under the coordination of multiple parameters.

[0003] A representative existing solution for this requirement is a multi-sensor feedback system based on Proportional Integral Derivative Control (PID). This solution deploys a temperature sensor array in different zones of the condensation tube wall to obtain annular temperature gradient data in real time, and combines a steam flow rate sensor and a condensation efficiency calculation model to construct a closed-loop feedback control loop. The system uses the PID algorithm to dynamically adjust the opening degree of the electric valve in the cooling water circulation system, and realizes the partition flow distribution through the preset rule of the baffle plate deflection angle. For example, when the temperature gradient in a certain zone exceeds the threshold, the PID controller will adjust the opening degree increment of the valve based on historical data, and at the same time, the baffle plate deflection angle will be gradually corrected according to the preset step size to improve the steam distribution uniformity.

[0004] However, the defects of the existing solution are mainly reflected in the insufficient adaptability of static rules to dynamic working conditions. First, the parameters of the PID controller rely on manual experience setting and cannot autonomously optimize the control strategy according to the dynamic correlation between the condensation efficiency and the phase change characteristics, resulting in a large energy efficiency loss due to a lag in response when the environmental temperature suddenly changes or the steam load fluctuates. Second, the adjustment rule of the baffle plate deflection angle is based on a fixed step size and threshold, lacking a quantitative model for the thermal inertia differences of multiple zones, and it is difficult to achieve a globally optimal thermal field balance. For example, when the thermal inertia coefficient differences of the condensation tube wall zones are large, the fixed-step adjustment may cause local overcooling or overheating, exacerbating energy loss. In addition, this system does not integrate a cooperative optimization mechanism for condensation efficiency and environmental temperature, resulting in a bottleneck in energy efficiency improvement. Summary of the Invention

[0005] The present application provides an intelligent data monitoring and analysis method and system for a solar desalination system, aiming to solve the problems in the prior art, including insufficient parameter self-optimization ability under dynamic working conditions, lack of quantitative modeling of multi-zone thermal inertia differences, and lack of a collaborative optimization mechanism between condensation efficiency and ambient temperature.

[0006] In a first aspect, the present application provides an intelligent data monitoring and analysis method for a solar desalination system, including: During the condensation stage of the solar desalination system, obtain the annular temperature gradient distribution data of the condensation tube wall, the phase change characteristic data of the steam in the condensation tube during condensation, and the ambient temperature data. Wherein, corresponding condensation structures are embedded in different zones inside the condensation tube wall, and each condensation structure includes a cooling water circulation system and an adjustable-angle deflector. Calculate the condensation efficiency based on the annular temperature gradient distribution data and the phase change characteristic data. Input the condensation efficiency and the ambient temperature data into a pre-trained reinforcement learning parameter optimization model, and combine the physical constraint conditions of the condensation structure to generate the flow regulation parameters and deflector deflection angle regulation parameters for each zone. Convert the flow regulation parameters and deflector deflection angle regulation parameters for each zone into corresponding electric control valve opening commands respectively, so as to synchronously adjust the opening of the control valve of the cooling water circulation system corresponding to different zones and the angle of the adjustable-angle deflector.

[0007] Optionally, the step of inputting the condensation efficiency and the ambient temperature data into a pre-trained reinforcement learning parameter optimization model, and combining the physical constraint conditions of the condensation structure to generate the flow regulation parameters and deflector deflection angle regulation parameters for each zone includes: Bind the condensation efficiency and the ambient temperature data in space-time according to the zone identification code, and input them into the feature extraction module of the reinforcement learning parameter optimization model to extract the dynamic feature sequence and the thermal inertia coefficient. Input the dynamic feature sequence into the dual-channel policy network of the reinforcement learning parameter optimization model. The first-channel policy network generates the initial flow parameters of the cooling water circulation system corresponding to each zone through a fully connected layer and a flow constraint layer. The second-channel policy network generates the initial deflection angle parameters of the adjustable-angle deflector corresponding to each zone through a convolutional layer and an angle constraint layer. The flow constraint layer dynamically adjusts the constraint boundary based on the flow constraint strategy output by the reinforcement learning agent, and the angle constraint layer dynamically adjusts the constraint boundary based on the angle constraint strategy output by the reinforcement learning agent. Take the initial flow parameter and the initial deflection angle parameter as the state input of the reinforcement learning agent, receive the parameter correction action output by the reinforcement learning agent, perform parameter correction based on the physical constraint conditions, and generate an intermediate flow parameter and an intermediate deflection angle parameter; Input the intermediate flow parameter and the intermediate deflection angle parameter into the multi-objective optimization module of the reinforcement learning parameter optimization model for parameter optimization; Perform partition-level cascade fusion on the optimized intermediate flow parameter, the optimized intermediate deflection angle parameter, and the historical optimal parameter, dynamically allocate fusion weights according to the thermal inertia coefficient of each partition condensation structure, and generate the flow rate adjustment parameter and the deflector deflection angle adjustment parameter for each partition.

[0008] Optionally, input the dynamic feature sequence into the dual-channel policy network of the reinforcement learning parameter optimization model, where the first-channel policy network generates the initial flow parameter of the cooling water circulation system corresponding to each partition through a fully connected layer and a flow constraint layer, and the second-channel policy network generates the initial deflection angle parameter of the adjustable angle deflector corresponding to each partition through a convolutional layer and an angle constraint layer, including: Input each partition feature vector in the dynamic feature sequence into the fully connected layer, and map it to a flow feature vector through dense connection between nodes; Input the flow feature vector into the flow constraint layer, where the maximum pressure threshold and the minimum flow rate threshold of the pipeline of the cooling water circulation system are predefined in the flow constraint layer, and perform element-wise truncation on the flow feature vector through an inequality constraint function to generate the initial flow parameter of the cooling water circulation system corresponding to each partition; Input the two-dimensional spatial distribution feature of each partition in the dynamic feature sequence into the convolutional layer, perform sliding window convolution along the partition spatial dimension through multiple groups of convolutional kernels, and extract the deflector deflection angle correlation feature between adjacent partitions; Input the deflector deflection angle correlation feature into the angle constraint layer, where the upper limit and the lower limit of the mechanical rotation angle of the adjustable angle deflector are predefined in the angle constraint layer, and map the deflector deflection angle correlation feature to a preset angle interval through a piecewise linear mapping function to generate the initial deflection angle parameter of the adjustable angle deflector corresponding to each partition.

[0009] Optionally, calculating the condensation efficiency based on the annular temperature gradient distribution data and the phase change feature data includes: Equally angularly divide the annular temperature gradient distribution data into a plurality of annular sub-regions in the radial direction of the annular region, and extract the temperature gradient vector data of each annular sub-region; Obtain the latent heat of phase change value and the upper limit value of the phase change temperature range for each annular sub-region from the phase change characteristic data, and calculate the heat response rate coefficient for each annular sub-region, where the heat response rate coefficient is the ratio of the latent heat of phase change value of the corresponding annular sub-region to the upper limit value of the phase change temperature range; Multiply the temperature gradient vector data of each annular sub-region by the heat response rate coefficient of the corresponding annular sub-region to generate the contribution value of the condensation efficiency for each annular sub-region; According to the area proportion of each annular sub-region in the annular region, perform weighted summation on the contribution value of the condensation efficiency of each annular sub-region to generate the condensation efficiency.

[0010] Optionally, input the flow characteristic vector into the flow constraint layer, where the maximum pressure threshold and the minimum flow rate threshold of the pipeline of the cooling water circulation system are predefined in the flow constraint layer, and perform element-by-element truncation on the flow characteristic vector through an inequality constraint function to generate the initial flow parameters of the cooling water circulation system corresponding to each partition, including: Perform element-by-element constraint processing on each element of the flow characteristic vector. If the current element value is greater than the maximum allowable flow value corresponding to the maximum pressure threshold of the pipeline in the corresponding partition, modify the current element value to the maximum allowable flow value; If the current element value is less than the minimum allowable flow value corresponding to the minimum flow rate threshold of the corresponding partition, modify the current element value to the minimum allowable flow value; If the current element value is between the minimum allowable flow value and the maximum allowable flow value, keep the current element value unchanged; Recombine the elements after element-by-element constraint processing according to the partition mapping relationship to generate the initial flow parameters of the cooling water circulation system corresponding to each partition.

[0011] Optionally, use the initial flow parameters and the initial deflection angle parameters as the state input of the reinforcement learning agent, receive the parameter correction actions output by the reinforcement learning agent, and perform parameter correction based on the physical constraint conditions to generate intermediate flow parameters and intermediate deflection angle parameters, including: Perform numerical comparison of the initial flow parameters with the upper limit value and the lower limit value of the constraint of the flow parameters respectively; Perform numerical comparison of the initial deflection angle parameters with the upper limit angle and the lower limit angle of the constraint of the deflection angle parameters respectively; Determine the intermediate flow parameters according to the flow comparison result, and determine the intermediate deflection angle parameters according to the deflection angle comparison result.

[0012] Optionally, the determining the intermediate flow parameters according to the flow comparison result and determining the intermediate deflection angle parameters according to the deflection angle comparison result includes: If the flow rate comparison result indicates that the initial flow rate parameter is greater than the upper limit value of the constraint of the flow rate parameter, the first fixed offset is used as the intermediate flow rate parameter; if the flow rate comparison result indicates that the initial flow rate parameter is less than the lower limit value of the constraint of the flow rate parameter, the second fixed offset is used as the intermediate flow rate parameter. If the deflection angle comparison result indicates that the initial deflection angle parameter is greater than the upper limit angle of the constraint of the deflection angle parameter, an intermediate deflection angle parameter is generated between the initial deflection angle parameter and the upper limit angle of the constraint of the deflection angle parameter through interpolation; if the initial deflection angle parameter is less than the lower limit angle of the constraint of the deflection angle parameter, an intermediate deflection angle parameter is generated between the lower limit angle of the constraint and the initial deflection angle parameter through interpolation.

[0013] In a second aspect, the present application provides an intelligent data monitoring and analysis system for a solar seawater desalination system, including: An acquisition module, configured to acquire the annular temperature gradient distribution data of the condensation tube wall, the phase change characteristic data of the steam in the condensation tube during the condensation process, and the ambient temperature data during the condensation stage of the solar seawater desalination system, wherein corresponding condensation structures are embedded in different partitions inside the condensation tube wall, and each of the condensation structures includes a cooling water circulation system and an adjustable angle deflector; A calculation module, configured to calculate the condensation efficiency based on the annular temperature gradient distribution data and the phase change characteristic data; A generation module, configured to input the condensation efficiency and the ambient temperature data into a pre-trained reinforcement learning parameter optimization model, and generate the flow rate adjustment parameter and the deflector deflection angle adjustment parameter for each partition in combination with the physical constraint conditions of the condensation structure; An adjustment module, configured to convert the flow rate adjustment parameter and the deflector deflection angle adjustment parameter for each partition into corresponding electric control valve opening commands respectively, so as to synchronously adjust the opening of the control valve of the cooling water circulation system corresponding to different partitions and the angle of the adjustable angle deflector.

[0014] In a third aspect, the present application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement an intelligent data monitoring and analysis method for a solar seawater desalination system as described in any item of the first aspect.

[0015] In a fourth aspect, the present application provides a computer storage medium, storing a computer program, and when the computer program is executed by a computer, it implements an intelligent data monitoring and analysis method for a solar seawater desalination system as described in any item of the first aspect.

[0016] In an embodiment of the present application, an intelligent data monitoring and analysis method for a solar seawater desalination system is provided. The method includes: during the condensation stage of the solar seawater desalination system, obtaining the annular temperature gradient distribution data of the condensation tube wall, the phase change characteristic data of the steam in the condensation tube during the condensation process, and the ambient temperature data, wherein corresponding condensation structures are embedded in different partitions inside the condensation tube wall, and each of the condensation structures includes a cooling water circulation system and an adjustable-angle deflector; calculating the condensation efficiency based on the annular temperature gradient distribution data and the phase change characteristic data; inputting the condensation efficiency and the ambient temperature data into a pre-trained reinforcement learning parameter optimization model, and generating the flow rate adjustment parameter and the deflector deflection angle adjustment parameter for each partition in combination with the physical constraint conditions of the condensation structure; and respectively converting the flow rate adjustment parameter and the deflector deflection angle adjustment parameter for each partition into corresponding electric control valve opening commands to synchronously adjust the opening of the control valve of the cooling water circulation system corresponding to different partitions and the angle of the adjustable-angle deflector.

[0017] Through the real-time monitoring of the multi-partition embedded condensation structure, the present application accurately captures the temperature field distribution of the condensation tube wall and the dynamic characteristics of steam phase change, solves the problem of inaccurate identification of local thermodynamic states caused by single temperature measurement points or global average measurements in traditional systems, and provides high-resolution data support for subsequent refined regulation. By combining the dynamic characteristics of the steam phase change process with the temperature gradient distribution, the present application realizes the quantitative evaluation of the partition-level condensation efficiency, breaks through the limitation of traditional methods relying on overall efficiency estimation, provides a scientific basis for partition-differentiated regulation, and avoids the drag of the overall system performance by local low-efficiency regions. By using the reinforcement learning model to fuse real-time environmental variables and physical constraints, the present application dynamically optimizes the flow rate and deflector deflection angle of each partition, solves the problem that traditional PID or rule-based control cannot adapt to non-linear heat load changes, and improves the adaptive ability of the system under variable working conditions. Through the coordinated action of the electric control valve and the deflector, the present application realizes the precise matching of multi-partition flow rate and air flow organization, overcomes the problems of uneven condensation or energy waste caused by traditional single-point regulation, and improves the energy efficiency and freshwater output stability of the system.

[0018] Furthermore, by spatiotemporally binding the condensation efficiency and ambient temperature data, the feature extraction module of the input reinforcement learning model extracts the dynamic feature sequence and the thermal inertia coefficient. The dual-channel policy network is used to separately process the flow rate and deflector parameter optimization. Specifically, the first channel generates the initial flow rate parameters through a fully connected layer and a dynamic flow rate constraint layer, and the second channel generates the initial deflector deflection angle parameters through a convolutional layer and an angle constraint layer. Both are based on the reinforcement learning agent to dynamically adjust the constraint boundaries. Subsequently, the initial parameters are subjected to constraint correction and multi-objective optimization, and finally, the historical optimal parameters and the thermal inertia coefficient are dynamically fused to output the optimal regulation parameters at the partition level. Among them, the flow rate constraint layer ensures the safe operation of the pipeline through an inequality function, the angle constraint layer ensures mechanical limits through a piecewise linear mapping, and the convolutional layer mines the spatial correlation features between adjacent partitions. Through dynamic feature extraction and dual-channel collaborative optimization, precise decoupled regulation of the flow rate and deflector parameters is achieved, solving the problem of response hysteresis caused by parameter coupling in traditional methods. The dynamic boundary adjustment of the constraint layer ensures the physical safety of the system and avoids the risk of overload. The parameter fusion based on the thermal inertia coefficient improves the adaptability to the nonlinear thermodynamic process, especially optimizing the regulation stability under variable working conditions. The introduction of spatial correlation features enhances the multi-partition collaborative control ability, effectively reducing the phenomenon of local condensation unevenness and overall improving the system energy efficiency and water production consistency.

[0019] These aspects or other aspects of the present application will be more clearly understood in the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0021] Figure 1 It is a flowchart of an intelligent data monitoring and analysis method for a solar desalination system provided by an embodiment of the present application; Figure 2 It is a schematic structural diagram of an intelligent data monitoring and analysis system for a solar desalination system provided by an embodiment of the present application; Figure 3 It is a schematic structural diagram of a computing device provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] In order to enable those skilled in the art to better understand the solutions of the present application, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application.

[0023] In some of the processes described in the specification, claims, and above-mentioned drawings of the present application, a number of operations appear in a specific order. However, it should be clearly understood that these operations may not be executed in the order in which they appear herein or may be executed in parallel. The operation numbers such as 11, 12, etc. are only used to distinguish different operations, and the numbers themselves do not represent any execution order. Additionally, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions such as "first", "second", etc. in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequence, nor do they limit that "first" and "second" are of different types.

[0024] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present application.

[0025] To solve the problems in the prior art, such as insufficient ability for autonomous optimization of parameters under dynamic working conditions, lack of quantitative modeling of multi-zone thermal inertia differences, and lack of a collaborative optimization mechanism for condensation efficiency and ambient temperature, the embodiments of the present application provide an intelligent data monitoring and analysis method for a solar desalination system. The method adopts the following concept: for the problem of optimizing the efficiency in the condensation stage of the solar desalination system, first, the annular temperature gradient distribution of the condensation tube wall, the steam phase change characteristics, and the ambient temperature data are collected in real time through a multi-zone embedded condensation structure to accurately perceive the local thermodynamic state; a partition condensation efficiency calculation model is established based on the temperature gradient and phase change data to quantitatively evaluate the performance differences in each region; subsequently, the condensation efficiency and ambient parameters are input into a pre-trained reinforcement learning optimization model, and combined with the physical constraint conditions of the cooling water circulation system and the deflector, the optimal control parameters for the flow rate in each zone and the deflection angle of the deflector are dynamically generated; finally, the parameters are converted into the opening commands of the electric control valves to achieve multi-zone synchronous and accurate adjustment of the cooling water flow rate and the guiding angle, thereby breaking through the bottleneck of lagging regulation and uneven energy efficiency in the traditional method under dynamic working conditions and improving the overall condensation efficiency and stability of the system.

[0026] Figure 1 The flowchart of an intelligent data monitoring and analysis method for a solar desalination system provided by the embodiments of the present application is as Figure 1 shown, and the method includes: S11. During the condensation stage of the solar desalination system, obtain the annular temperature gradient distribution data of the condensation tube wall, the phase change characteristic data of the steam in the condensation tube during the condensation process, and the ambient temperature data. The condensation structures corresponding to different partitions are embedded inside the condensation tube wall, and each condensation structure includes a cooling water circulation system and an adjustable-angle deflector.

[0027] Among them, the annular temperature gradient distribution data refers to the temperature values of each temperature measurement point along the circumferential direction of the condensation tube and their spatial change rates. The phase change characteristic data includes physical quantities during the phase change process such as the steam condensation rate, liquid film thickness, and latent heat release amount. The ambient temperature data refers to the real-time air temperature of the space where the condensation device is located during the operation of the solar desalination system. The data source is collected through anti-corrosion temperature sensors or multi-point infrared temperature measurement modules arranged on the outer wall of the condensation tube, and the sampling frequency ≥ 1 Hz. The cooling water circulation system can refer to an active heat exchange subsystem integrated in each condensation partition. The adjustable-angle deflector refers to a metal deflector blade that can rotate 0 - 90° around an axis and is used to regulate the direction of the condensation airflow.

[0028] In the embodiment of the present application, during the condensation stage of the solar desalination system, the annular temperature gradient distribution data is obtained through a temperature sensor array arranged in each partition of the condensation tube wall. At the same time, a steam phase change monitoring device is used to collect the phase change characteristic data of the steam in the condensation tube, and the ambient temperature data is recorded synchronously. Among them, the condensation tube wall adopts a partition-embedded design, and each partition integrates an independent condensation structure, which includes a cooling water circulation system and an adjustable-angle deflector. When collecting data, the temperature sensors are arranged in an annular array and sampled once every 5 seconds. The steam phase change data can be obtained by real-time capture using a high-precision optical sensor.

[0029] S12. Calculate the condensation efficiency based on the annular temperature gradient distribution data and the phase change characteristic data.

[0030] Among them, the condensation efficiency refers to the ratio of the freshwater energy actually condensed to the latent heat of the input steam.

[0031] In the embodiment of the present application, first, the annular temperature gradient data is averaged and the standard deviation is calculated for each partition to characterize the uniformity of the temperature field. Then, the condensation rate in the steam phase change characteristic data is divided by the theoretical maximum condensation rate of the partition to obtain the phase change efficiency coefficient. Finally, the temperature field uniformity index is multiplied by the phase change efficiency coefficient and then multiplied by a normalized weight factor to output the partition condensation efficiency value within the range of 0 - 1. Optionally, to improve the calculation efficiency, the calculation process for each partition can be executed in parallel.

[0032] S13. Input the condensation efficiency and the ambient temperature data into a pre-trained reinforcement learning parameter optimization model, and combine the physical constraint conditions of the condensation structure to generate the flow rate adjustment parameter and the deflector deflection angle adjustment parameter for each partition.

[0033] Among them, the reinforcement learning parameter optimization model is an intelligent decision-making system based on the proximal policy optimization algorithm, which is specifically used to dynamically optimize the condensation control parameters of the solar desalination system. The physical constraint conditions refer to the maximum / minimum flow rate, pressure, angle and other limit parameters allowed by the mechanical structure of the system. The flow rate adjustment parameter refers to the optimization instruction output by the reinforcement learning model for dynamically adjusting the flow rate of the cooling water circulation system. The deflector deflection angle adjustment parameter refers to the instruction output by the reinforcement learning model for optimizing the position of the adjustable angle deflector. Therefore, S13 can input the condensation efficiency and the environmental temperature data into the reinforcement learning parameter optimization model pre-trained based on the proximal policy optimization algorithm, and combine the physical constraint conditions of the condensation structure to generate the flow rate adjustment parameters and the deflector deflection angle adjustment parameters for each partition; the input dimension of the reinforcement learning parameter optimization model is the joint feature vector of the condensation efficiency, the environmental temperature and the partition identification code, and the output dimension is the flow rate parameter and the deflection angle parameter for each partition.

[0034] In the embodiment of the present application, after encoding the condensation efficiency and the environmental temperature data by partition, they are input into the pre-trained reinforcement learning parameter optimization model. The model first verifies the validity of the input data through the physical constraint condition check module, and the constraint conditions include the maximum flow rate limit of the cooling water circulation system and the mechanical angle range of the deflector. The double-branch output layer of the model respectively generates the flow rate adjustment parameter and the deflector deflection angle adjustment parameter. Among them, the flow rate parameter is mapped to the constraint range through the s function, and the deflector parameter is corrected by piecewise linear interpolation. Optionally, the parameter matrix can be updated once every 30 seconds in the embodiment of the present application.

[0035] S14. Convert the flow rate adjustment parameter and the deflector deflection angle adjustment parameter for each partition into corresponding electric control valve opening instructions respectively, so as to synchronously adjust the opening degree of the control valve of the cooling water circulation system corresponding to different partitions and the angle of the adjustable angle deflector.

[0036] Among them, the electric control valve opening instruction refers to the standardized electric signal for controlling the opening degree of the valve.

[0037] In the embodiment of the present application, the flow rate adjustment parameter for each partition is converted into an analog signal to drive the opening of the electric control valve. At the same time, the deflector deflection angle parameter controls the rotation of the servo motor through a Pulse Width Modulation (PWM) signal. The adjustment process can adopt the PID synchronization algorithm to ensure that all partitions are adjusted within 2 seconds. The change rate of the cooling water flow rate is limited to ±10% / s, and the angular velocity of the deflector does not exceed 5° / s to avoid hydraulic shock or mechanical vibration. The system real-time feedbacks the valve opening degree and the deflector deflection angle to the central controller to form a closed loop.

[0038] The following is a specific example: After a certain solar desalination plant deploys this system, during the strong sunlight at noon, the annular temperature gradient in Zone 3 is detected to be abnormal, and its condensation efficiency drops to 0.65; based on this, the reinforcement learning model generates parameters for increasing the flow rate in this zone by 8.2 L / min and adjusting the deflector to 52°; the efficiency is restored to 0.82 within 2 seconds after execution. During the same period, due to the uniform temperature gradient caused by the shadow in Zone 7, the system automatically reduces its flow rate by 5% and increases the deflection angle of the deflector to 65°, so that the overall water production rate remains stable at 8.2 L / (m²·h).

[0039] By executing S11~S14, through multi-source data fusion and dynamic optimization of reinforcement learning, precise control of each zone in the condensation process is achieved: the collaborative analysis of temperature gradient and phase change data improves the accuracy of efficiency calculation; the parameter generation of the reinforcement learning model under physical constraints ensures the safety of the system and avoids overload faults; the synchronous control of the electric valve and the deflector shortens the working condition adjustment time and also improves the steam utilization rate.

[0040] Therefore, through the real-time monitoring of the multi-zone embedded condensation structure in the embodiment of the present application, the temperature field distribution of the condensation tube wall and the dynamic characteristics of steam phase change are accurately captured, solving the problem of inaccurate identification of local thermodynamic states caused by single temperature measurement points or global average measurements in traditional systems, and providing high-resolution data support for subsequent refined control. Combining the dynamic characteristics of the steam phase change process and the temperature gradient distribution, a quantitative evaluation of the condensation efficiency at the zone level is realized, breaking through the limitation of traditional methods relying on overall efficiency estimation, providing a scientific basis for zone-level differential control, and avoiding the overall system performance being dragged down by local low-efficiency areas. Using the reinforcement learning model to fuse real-time environmental variables and physical constraints, dynamically optimize the flow rate and deflector deflection angle parameters of each zone, solve the problem that traditional PID or rule control cannot adapt to non-linear heat load changes, and improve the adaptive ability of the system under variable working conditions. Through the coordinated action of the electric control valve and the deflector, precise matching of the flow rate and air flow organization in multiple zones is achieved, overcoming the problems of uneven condensation or energy waste caused by traditional single-point control, and improving the energy efficiency of the system and the stability of fresh water output.

[0041] In a possible embodiment, in S13, input the condensation efficiency and environmental temperature data into a pre-trained reinforcement learning parameter optimization model, and combine the physical constraint conditions of the condensation structure to generate the flow rate adjustment parameters and deflector deflection angle adjustment parameters for each zone, including: Step 131: Bind the condensation efficiency and environmental temperature data in space-time according to the zone identification code, and input them into the feature extraction module of the reinforcement learning parameter optimization model to extract the dynamic feature sequence and the thermal inertia coefficient.

[0042] Among them, spatio-temporal binding refers to associatively encoding data in different partitions according to timestamps and spatial positions to form a data matrix with spatio-temporal attributes. The dynamic feature sequence refers to a multi-dimensional time-series feature set extracted from the condensation efficiency and ambient temperature data of spatio-temporal binding, which reflects the real-time operating state of the system. The thermal inertia coefficient refers to a dimensionless parameter that measures the thermal response delay of a partition.

[0043] In the embodiment of the present application, first, the condensation efficiency and ambient temperature data are spatio-temporally bound according to the partition identifier encoding and input into the feature extraction module of the reinforcement learning parameter optimization model. The feature extraction module can adopt a hybrid structure of a convolutional neural network and a long short-term memory network, extract the local spatial correlation of temperature gradients and phase change features through the convolutional layer, and capture the time-series dependence of ambient temperature changes through the long short-term memory network layer. Finally, the module outputs a dynamic feature sequence and a thermal inertia coefficient.

[0044] In other embodiments, the product of the time interval and the heat flow is calculated, and the ratio of the partition temperature change amount to the product result is used as the thermal inertia coefficient. The constraint boundaries of the flow constraint layer and the angle constraint layer are dynamically adjusted based on the real-time feedback of the pipeline pressure sensor data and the mechanical limit data of the deflector angle sensor.

[0045] Step 132: Input the dynamic feature sequence into the dual-channel policy network of the reinforcement learning parameter optimization model. The first-channel policy network generates the initial flow parameters of the cooling water circulation system corresponding to each partition through a fully connected layer and a flow constraint layer. The second-channel policy network generates the initial deflection angle parameters of the adjustable angle deflector corresponding to each partition through a convolutional layer and an angle constraint layer. The flow constraint layer dynamically adjusts the constraint boundary based on the flow constraint strategy output by the reinforcement learning agent, and the angle constraint layer dynamically adjusts the constraint boundary based on the angle constraint strategy output by the reinforcement learning agent.

[0046] Among them, the dual-channel policy network is a neural network architecture that processes different regulation objectives in parallel, and generates flow parameters and deflector angle adjustment parameters through two independent channels respectively. The fully connected layer is a neural network layer whose nodes are densely connected to all nodes in the previous layer, and is used for global feature integration and non-linear mapping. The flow constraint layer is a neural network layer that ensures that the flow does not exceed the pipe pressure bearing and flow velocity limits. The initial flow parameter is the preliminary regulation value output by the first channel in the dual-channel policy network, representing the recommended flow distribution of the cooling water circulation system. The convolutional layer extracts local spatial features through the sliding window operation of the convolutional kernel, and is suitable for processing data with spatial correlation. The angle constraint layer is a logic module that limits the deflector angle within the mechanically feasible range. The initial deflection angle parameter is the preliminary angle recommendation value output by the second channel in the dual-channel policy network, representing the recommended deflection angles of the deflectors in each partition. The reinforcement learning agent can be an intelligent decision-making module using the deep Q-network or proximal policy optimization algorithm, and outputs parameter adjustment actions.

[0047] In the embodiment of the present application, the dynamic feature sequence is input into the dual-channel policy network. The first-channel policy network maps the feature vector to a high-dimensional space through the fully connected layer to generate the initial value of the flow parameter; the flow constraint layer dynamically adjusts the parameter boundary through the inequality truncation function based on the flow constraint policy output by the reinforcement learning agent, and generates the initial flow parameters of the cooling water circulation system in each partition. The second-channel policy network extracts the associated features of the deflector angles of adjacent partitions through the convolutional layer, and the angle constraint layer converts the features to a preset interval using the piecewise linear mapping function according to the upper and lower limits of the mechanical rotation angle, and generates the initial deflection angle parameters of the adjustable angle deflectors in each partition. The dual-channel design realizes the decoupled optimization of flow and angle. It should be noted that the embodiments of the present application do not specifically limit the expressions corresponding to the inequality truncation function and the piecewise linear mapping function.

[0048] Step 133: Use the initial flow parameter and the initial deflection angle parameter as the state input of the reinforcement learning agent, receive the parameter correction actions output by the reinforcement learning agent, perform parameter correction based on the physical constraint conditions, and generate the intermediate flow parameter and the intermediate deflection angle parameter.

[0049] Among them, the parameter correction action refers to the flow / angle adjustment instruction output by the reinforcement learning agent. The intermediate flow parameter and the intermediate deflection angle parameter are the correction results of the reinforcement learning agent for the initial flow parameter and the initial deflection angle parameter.

[0050] In the embodiment of the present application, the initial flow rate parameter and the initial deflection angle parameter are used as the state inputs of the reinforcement learning agent. The agent adopts the common Deep Deterministic Policy Gradient algorithm to correct actions by outputting parameters and evaluate the action values. The corrected actions need to satisfy the physical constraint conditions, and the corrected parameters are restricted within the feasible region through the common constraint projection algorithm, and finally the intermediate flow rate parameter and the intermediate deflection angle parameter are generated. This process can achieve the balance between local optimality and global constraints.

[0051] Step 134: Input the intermediate flow rate parameter and the intermediate deflection angle parameter into the multi-objective optimization module of the reinforcement learning parameter optimization model for parameter optimization. Among them, the multi-objective optimization module refers to the optimization algorithm component that balances efficiency, energy consumption, and safety.

[0052] In the embodiment of the present application, the intermediate flow rate parameter and the intermediate deflection angle parameter are input into the multi-objective optimization module. Based on the Pareto optimization framework, the module aims to maximize the condensation efficiency, minimize the energy consumption, and achieve the thermal field balance. The expression of this optimization objective is a conventional mathematical expression and will not be elaborated here. The non-dominated sorting genetic algorithm is used to generate the parameter optimization solution set. During the optimization process, the module dynamically adjusts the objective weights in combination with the real-time fluctuation characteristics of the ambient temperature data and outputs the optimized parameters with multi-objective coordination.

[0053] Step 135: Perform partition-level cascade fusion on the optimized intermediate flow rate parameter, the optimized intermediate deflection angle parameter, and the historical optimal parameters, and dynamically allocate the fusion weights according to the thermal inertia coefficients of the condensation structures in each partition to generate the flow rate adjustment parameter and the deflector deflection angle adjustment parameter for each partition.

[0054] Among them, the partition-level cascade fusion is the process of weighted fusion of real-time parameters and historical parameters according to the partition characteristics.

[0055] In the embodiment of the present application, the optimized intermediate parameters and the historical optimal parameters are subjected to partition-level cascade fusion. The fusion process can be realized by using the weighted average algorithm, and the weights are dynamically allocated by the thermal inertia coefficients of each partition. For example, for the partition with obvious heat conduction lag, its flow rate adjustment parameter is mainly based on the historical optimal value, while for the partition with fast thermal response, it is mainly based on the real-time optimized value. The finally generated flow rate adjustment parameter and the deflector deflection angle adjustment parameter are mapped to the actuator through partition coding to achieve global dynamic regulation.

[0056] The following is a specific example: During the noon operation period of a certain solar desalination plant, the system detected an abnormal annular temperature gradient in Zone 3 and the condensation efficiency dropped to 0.65. At the same time, Zone 7 showed a uniform temperature of 28 °C due to shadow coverage. The system bound the efficiency data of 0.65 in Zone 3 with the temperature gradient data according to the zone code, extracted the dynamic feature sequence through the network and calculated the thermal inertia coefficient of 1.1; after synchronously binding the data in Zone 7, the feature sequence and the thermal inertia coefficient of 0.9 were extracted. The features were input into the dual-channel policy network: The first channel generated the initial flow rate parameter for Zone 3, which was dynamically limited to [+5 L / min, +10 L / min] by the flow rate constraint layer and then output +8.2 L / min; the second channel generated the initial deflection angle of the deflector of 50°, which was limited to [45°, 60°] by the angle constraint layer and then output 52°; Zone 7 generated a 6% reduction in flow rate and an angle of 63°. The reinforcement learning agent fine-tuned the parameters of Zone 3 based on physical constraints, keeping +8.2 L / min and 52° unchanged. Zone 7 was finally confirmed to have a -5% flow rate and an angle of 65°. Through the multi-objective optimization module to balance efficiency and energy consumption, the parameters of Zone 3 were optimized to +8.0 L / min and 53°, and Zone 7 was adjusted to -4.8% and 64°. The optimized parameters were weighted and fused with the historical optimal values according to the thermal inertia coefficient. Finally, the flow rate of Zone 3 was output as +7.7 L / min and the deflector angle was 51.5°, and the flow rate of Zone 7 was -4.9% and the angle was 62°, enabling the efficiency of Zone 3 to recover to 0.82 within 2 seconds and the water production rate of the entire solar desalination plant to stabilize at 8.2 L / (m²·h).

[0057] By performing Steps 131 to 135, the embodiment of the present application accurately depicts the system state through spatio-temporal binding and dynamic feature extraction, optimizes the flow rate and the deflector deflection angle respectively by combining with the dual-channel policy network, uses the reinforcement learning agent and the multi-objective optimization module to realize automatic parameter correction and global optimization, and finally dynamically weights and fuses the historical data through the thermal inertia coefficient to improve the adaptive ability and energy efficiency ratio of the condensation system, while reducing equipment loss.

[0058] In a possible embodiment, in Step 132, the dynamic feature sequence is input into the dual-channel policy network of the reinforcement learning parameter optimization model. The first-channel policy network generates the initial flow rate parameter of the cooling water circulation system corresponding to each zone through the fully connected layer and the flow rate constraint layer, and the second-channel policy network generates the initial deflection angle parameter of the adjustable-angle deflector corresponding to each zone through the convolutional layer and the angle constraint layer, including: Step a1: Input each zone feature vector in the dynamic feature sequence into the fully connected layer, and map it to a flow rate feature vector through dense connection between nodes. Among them, the flow rate feature vector is the flow rate regulation-related feature encoded by the fully connected layer.

[0059] In the embodiment of the present application, the feature vectors of each partition in the dynamic feature sequence are input into a fully connected layer, which performs non-linear transformation through dense connections between nodes, and uses an activation function to map the features into low-dimensional flow feature vectors. This vector implies key factors such as pressure, flow velocity, heat exchange rate, etc. required for flow regulation, providing a basis for subsequent flow constraints.

[0060] Step a2: Input the flow feature vector into the flow constraint layer. The maximum pressure-bearing threshold and minimum flow velocity threshold of the pipeline of the cooling water circulation system are predefined in the flow constraint layer, and the flow feature vector is truncated element by element through an inequality constraint function to generate the initial flow parameters of the cooling water circulation system corresponding to each partition.

[0061] Among them, the formula of the inequality constraint function is , where is the element of the i-th partition of the flow feature vector, is the flow upper limit threshold corresponding to the maximum pressure-bearing of the pipeline, is the minimum flow velocity threshold allowed by the system, and is the initial flow parameter corresponding to the i-th partition. Truncation refers to the operation of forcing the flow recommendation value outside the physically feasible range to the preset boundary. Upward truncation: If the initial parameter exceeds , then force it to take . Downward truncation: If the initial parameter is lower than , then force it to take .

[0062] In the embodiment of the present application, the flow feature vector is input into the flow constraint layer, which predefines physical constraint conditions: the maximum pressure-bearing threshold and minimum flow velocity threshold of the pipeline. The flow feature vector is amplitude-limited element by element through an inequality constraint function to generate practical initial flow parameters.

[0063] Step a3: Input the two-dimensional spatial distribution features of each partition in the dynamic feature sequence into the convolutional layer, and perform sliding window convolution along the partition spatial dimension through multiple groups of convolutional kernels to extract the associated features of the deflector deflection angle between adjacent partitions.

[0064] Among them, the two-dimensional spatial distribution feature refers to the thermal map or distribution matrix of parameters such as the partition temperature field and deflector deflection angle on the two-dimensional plane. The partition spatial dimension refers to the physical space arrangement mode of the condensation partition in the solar seawater desalination plant, and each partition has a fixed spatial coordinate. Sliding window convolution refers to that the convolutional kernel slides at a fixed step length along the partition spatial dimension and calculates the weighted sum of local region features at each position. The associated features of the deflector deflection angle can include: air flow interference coefficient, temperature gradient propagation factor, efficiency coupling weight, mechanical stress distribution feature, etc.

[0065] In the embodiment of the present application, the two-dimensional spatial distribution features in the dynamic feature sequence are input into the convolutional layer, and a 3×3 convolutional kernel is used to slide the window along the partition space to calculate and extract the correlation features of the deflector deflection angles between adjacent partitions. Multiple groups of convolutional kernels extract different spatial patterns in parallel and output a fused angle correlation feature tensor.

[0066] Step a4: Input the deflector deflection angle correlation features into the angle constraint layer. The mechanical rotation angle upper and lower limits of the adjustable angle deflector are predefined in the angle constraint layer, and the deflector deflection angle correlation features are mapped to a preset angle interval through a piecewise linear mapping function to generate the initial deflection angle parameters of the adjustable angle deflector corresponding to each partition.

[0067] Among them, the formula of the piecewise linear mapping function is , where is the original angle suggestion value output by the convolutional layer, is the initial deflection angle parameter, is the mechanical rotation lower limit of the deflector, is the mechanical rotation upper limit of the deflector, is the first piecewise threshold, is the second piecewise threshold, is the third piecewise threshold, is the rising segment slope, is the falling segment slope, where , and The numerical values of can be constants.

[0068] In the embodiment of the present application, the deflector deflection angle correlation features are input into the angle constraint layer, and the mechanical limit is predefined in this layer. The features are mapped to a preset interval through a piecewise linear mapping function to generate the initial deflection angle parameters.

[0069] The following is a specific example: In a solar desalination plant, the 12-dimensional dynamic features of partition No. 3 are input into the fully connected layer, and an 8-dimensional flow feature vector is output. The vector element [1.1] corresponding to the flow +9 L / min is truncated to +8.2 L / min through the constraint layer. The temperature field matrix is convolved to extract the air flow interference feature 0.7 between partition No. 3 and the adjacent partition No. 4. This feature is mapped to 52°, and finally, a regulation instruction for the flow of partition No. 3 of +8.2 L / min and the deflector of 52° is generated, and the efficiency is restored to 0.82 within 2 seconds.

[0070] By performing steps a1 to a4, the embodiments of the present application extract flow regulation features and the associated features of the deflector deflection angle through a fully connected layer and a convolutional layer respectively. Combining with the dynamic boundary restriction of the physical constraint layer, it ensures that the generated regulation parameters always meet the system's mechanical bearing capacity and operation safety requirements. This method realizes the collaborative optimization of multi-zone parameters. On the premise of ensuring the safe operation of the equipment, through feature-driven intelligent adjustment, the flow rate and deflector deflection angle of each zone reach the optimal matching, effectively improving the overall stability and regulation response speed of the system, and at the same time avoiding air flow interference or pressure fluctuations between zones.

[0071] In a possible embodiment, S12, calculating the condensation efficiency based on the annular temperature gradient distribution data and the phase change characteristic data, includes: Step 121, equally angularly dividing the annular temperature gradient distribution data in the radial direction of the annular region into a plurality of annular sub-regions, and extracting the temperature gradient vector data of each annular sub-region.

[0072] Among them, the annular sub-region is a fan-shaped partition divided by a fixed angle for local temperature analysis. The temperature gradient vector data is a vector containing the temperature change direction and intensity.

[0073] In the embodiments of the present application, the annular temperature gradient distribution data is equally angularly divided in the radial direction of the annular region into a plurality of annular sub-regions. Through the temperature field interpolation algorithm in the polar coordinate system, the temperature gradient vector data of each annular sub-region is extracted, including the temperature change direction and intensity to form vector data.

[0074] Step 122, obtaining the latent heat of phase change value and the upper limit value of the phase change temperature range of each annular sub-region from the phase change characteristic data, and calculating the heat response rate coefficient of each annular sub-region. The heat response rate coefficient is the ratio of the latent heat of phase change value of the corresponding annular sub-region to the upper limit value of the phase change temperature range.

[0075] Among them, the latent heat of phase change value refers to the latent heat absorbed / released during the phase change of a substance. The upper limit value of the phase change temperature range refers to the highest critical temperature at which the material undergoes a phase change. The heat response rate coefficient is the ratio of the latent heat of phase change value to the temperature upper limit, reflecting the heat response sensitivity.

[0076] In the embodiments of the present application, the latent heat of phase change value and the upper limit value of the phase change temperature range of each annular sub-region are obtained from the phase change characteristic database. When calculating the heat response rate coefficient, the latent heat of phase change value is divided by the upper limit value of the phase change temperature range. This coefficient characterizes the ability of the material to absorb / release heat under unit temperature change and is used to quantify the contribution potential of the sub-region to the condensation efficiency.

[0077] Step 123, multiplying the temperature gradient vector data of each annular sub-region by the heat response rate coefficient of the corresponding annular sub-region to generate the condensation efficiency contribution value of each annular sub-region.

[0078] Among them, the contribution value of the condensation efficiency refers to the quantitative contribution of the sub-region to the overall condensation efficiency. The contribution value of the condensation efficiency refers to the quantitative contribution of each annular sub-region to the overall condensation efficiency based on its temperature gradient and thermal response characteristics.

[0079] In the embodiment of the present application, the temperature gradient vector data of each annular sub-region is subjected to a scalar multiplication operation with the corresponding thermal response rate coefficient to generate a condensation efficiency contribution value. This value comprehensively reflects the heat transfer rate driven by the temperature gradient and the thermal response characteristics of the phase change material.

[0080] Step 124: According to the area ratio of each annular sub-region in the annular region, perform a weighted sum of the condensation efficiency contribution values of each annular sub-region to generate the condensation efficiency.

[0081] Among them, the area ratio refers to the ratio of the sub-region area to the total area of the annular region.

[0082] In the embodiment of the present application, according to the area ratio of each annular sub-region, a weighted sum of the condensation efficiency contribution values is performed. Specifically: multiply the contribution value of each sub-region by its area weight, and then accumulate the results of all sub-regions to finally generate the overall condensation efficiency value. This process ensures the dominant influence of large-sized sub-regions on the efficiency through area weighting.

[0083] The following is a specific example: In the condensation ring of a certain solar desalination plant, an annular region with a diameter of 10 m is equally divided into 12 sub-regions of 30°. The temperature gradient vector [45°, 4.2 °C / m] of sub-region 3 is extracted; the latent heat of phase change value of this sub-region is 230 kJ / kg and the upper temperature limit is 75 °C, and the thermal response rate coefficient 3.07 kJ / (kg·°C) is calculated; a condensation efficiency contribution value of 4.2×3.07 = 12.89 kJ / (kg·m) is generated; according to the area ratio of sub-region 3 of 8.3%, the weighted value 12.89×0.083 = 1.07 is calculated, and finally the weighted sum of 12 sub-regions is 0.85, accurately reflecting the actual condensation efficiency of the system.

[0084] By performing steps 121 to 124, the embodiment of the present application accurately quantifies the contribution of each local region to the condensation efficiency through the refined segmentation of the annular sub-region and the coupled calculation of the thermal response, and realizes the scientific evaluation of the overall efficiency in combination with the area weight, solves the problem of insufficient correlation between the temperature gradient and the phase change characteristics in the traditional method, and improves the spatial resolution and accuracy of the condensation efficiency measurement.

[0085] In a possible embodiment, in step a2, the flow feature vector is input into a flow constraint layer. In the flow constraint layer, the maximum pressure-bearing threshold and the minimum flow velocity threshold of the pipeline of the cooling water circulation system are predefined. The flow feature vector is truncated element by element through an inequality constraint function to generate the initial flow parameters of the cooling water circulation system corresponding to each partition, including: Step a21: Perform element-by-element constraint processing on each element of the flow feature vector. If the current element value is greater than the maximum allowable flow value corresponding to the maximum pressure-bearing threshold of the pipeline in the corresponding partition, the current element value is modified to the maximum allowable flow value.

[0086] Among them, the maximum pressure-bearing threshold of the pipeline is the pressure-bearing limit value calculated from the pipeline material, wall thickness, safety factor, etc. The maximum allowable flow value is the flow upper limit converted based on the maximum pressure-bearing, and is related to the pipeline cross-sectional area and flow velocity.

[0087] Step a22: If the current element value is less than the minimum allowable flow value corresponding to the minimum flow velocity threshold of the corresponding partition, the current element value is modified to the minimum allowable flow value.

[0088] Among them, the minimum flow velocity threshold refers to the minimum flow velocity requirement to prevent siltation or freezing, and the minimum allowable flow value refers to the flow lower limit calculated from the minimum flow velocity and the pipeline cross-sectional area.

[0089] Step a23: If the current element value is between the minimum allowable flow value and the maximum allowable flow value, the current element value remains unchanged. Among them, the interval judgment is through a conditional judgment algorithm to confirm that the current flow value is between the minimum allowable flow value and the maximum allowable flow value. The conditional judgment algorithm is a prior art and will not be elaborated here.

[0090] Step a24: Recombine the elements after the element-by-element constraint processing according to the partition mapping relationship to generate the initial flow parameters of the cooling water circulation system corresponding to each partition. Among them, the partition mapping relationship is the corresponding rule between the partition number and the physical / logical area.

[0091] The following is a specific example: In a solar seawater desalination plant, the flow characteristic value of +11 L / min in the 3rd partition is truncated to +10 L / min; the recommended value of -6% in the 7th partition is increased to -5%; the reasonable value of +3.2 L / min in the 5th partition is retained; and they are recombined into a vector [+10 L / min, +3.2 L / min, -5%] according to the partition coding [3, 5, 7] to generate the initial flow parameters and send them to the circulation pump control system to ensure that the 3rd partition is not over-pressurized, the 7th partition is not silted, and the 5th partition is accurately regulated.

[0092] By performing steps a21 to a24, the embodiments of the present application convert the flow regulation recommended value output by the neural network into physically feasible initial parameters through partition - adaptive element - by - element constraint processing. On the premise of ensuring pipeline safety and system stability, it not only avoids the risks of overpressure or insufficient flow, but also maximally retains the optimization intention of the intelligent model, realizing the unity of safety and regulation effect.

[0093] In a possible embodiment, step 133: Use the initial flow parameters and the initial deflection angle parameters as the state inputs of the reinforcement learning agent, receive the parameter correction actions output by the reinforcement learning agent, and perform parameter correction based on physical constraint conditions to generate intermediate flow parameters and intermediate deflection angle parameters, including: Step b1: Numerically compare the initial flow parameters with the upper limit value and the lower limit value of the flow parameter constraints respectively.

[0094] Among them, the upper limit value of the constraint is the maximum allowable flow value calculated according to parameters such as pipeline material and design pressure. The lower limit value of the constraint is the minimum allowable flow value calculated based on the minimum flow velocity threshold. The numerical comparison is a process of comparing the parameters with the preset threshold through an algorithm and making corrections.

[0095] Step b2: Numerically compare the initial deflection angle parameters with the upper limit angle and the lower limit angle of the deflection angle parameter constraints respectively.

[0096] Among them, the upper limit angle of the constraint refers to the maximum allowable deflection angle determined by mechanical structure limits or fluid dynamics requirements. The lower limit angle of the constraint refers to the minimum deflection angle required to maintain the basic functions of the system.

[0097] Step b3: Determine the intermediate flow parameters according to the flow comparison result, and determine the intermediate deflection angle parameters according to the deflection angle comparison result.

[0098] Among them, the flow comparison result refers to the set of flow parameters after correction, ensuring that the flow in each partition is within a safe range. The deflection angle comparison result refers to the set of corrected deflection angle parameters, meeting mechanical and fluid dynamics constraints.

[0099] The following is a specific example: During the operation of a solar desalination plant, it is found that the initial flow of zone 7 is - 6%, which is lower than the lower limit of - 5%, and is marked as "needs to be increased"; it is verified that the deflector of zone 3 is at 52°, within the range of 20° to 70°, and is marked as "compliant"; the flow of zone 7 is corrected to - 5%, and the parameters of zone 3 remain unchanged. Finally, the intermediate parameters are output: flow [+8.2 L / min, - 5%], deflector deflection angle 65°, which not only meets the pipeline safety requirements but also maintains the regulation accuracy.

[0100] By performing steps b1 to b3, the embodiments of the present application adopt a dual-parameter boundary verification mechanism to ensure that all intermediate parameters strictly comply with the physical system constraints while retaining the intelligent regulation intention, providing a safe and reliable input for subsequent optimization, and effectively avoiding equipment damage or efficiency reduction caused by parameter overrun.

[0101] In a possible embodiment, step b3, determining the intermediate flow parameter according to the flow comparison result and determining the intermediate deflection angle parameter according to the deflection angle comparison result, includes: Step c1, if the flow comparison result indicates that the initial flow parameter is greater than the upper limit value of the constraint of the flow parameter, then use the first fixed offset as the intermediate flow parameter; if the flow comparison result indicates that the initial flow parameter is less than the lower limit value of the constraint of the flow parameter, then use the second fixed offset as the intermediate flow parameter.

[0102] Wherein, the first fixed offset refers to the safe downward adjustment amount when the flow exceeds the upper limit, which is used to prevent sudden pressure drop. The second fixed offset refers to the safe upward adjustment amount when the flow is lower than the lower limit, which is used to prevent sudden change in flow velocity.

[0103] Step c2, if the deflection angle comparison result indicates that the initial deflection angle parameter is greater than the upper limit angle of the constraint of the deflection angle parameter, then generate an intermediate deflection angle parameter between the initial deflection angle parameter and the upper limit angle of the constraint of the deflection angle parameter through interpolation operation; if the initial deflection angle parameter is less than the lower limit angle of the constraint of the deflection angle parameter, then generate an intermediate deflection angle parameter between the lower limit angle and the initial deflection angle parameter through interpolation operation.

[0104] Wherein, the interpolation operation is a mathematical method for calculating the transition value between the boundary value and the initial value in proportion. The interpolation coefficient is a weight factor that determines whether the intermediate parameter is closer to the boundary or the initial value.

[0105] The following is a specific example: In a solar desalination plant, apply the first offset of -1.5 L / min to the over-limit flow rate of +11 L / min in Zone 3 (the upper limit is +10 L / min), and output an intermediate value of 8.5 L / min; for the out-of-bounds angle of 75° (the upper limit is 70°) in Zone 5, generate 72° according to the interpolation coefficient of 0.6. Finally, the intermediate parameters are flow rate [8.5 L / min, -4.2%] and angle [72°, 18°], realizing a smooth transition from the out-of-bounds state to the safe range.

[0106] By performing steps c1 to c2, the embodiments of the present application generate a smoothly transitioning intermediate value through the combined application of the offset and interpolation algorithms, which not only satisfies the physical constraints but also retains the regulation continuity, effectively suppressing system oscillations caused by parameter jumps and improving control stability and equipment life.

[0107] Figure 2The structural schematic diagram of an intelligent data monitoring and analysis system for a solar seawater desalination system provided by an embodiment of the present application is as follows Figure 2 As shown, the system includes: An acquisition module 21, configured to obtain the annular temperature gradient distribution data of the condensation tube wall, the phase change characteristic data of the steam in the condensation tube during the condensation process, and the ambient temperature data during the condensation stage of the solar seawater desalination system. Corresponding condensation structures are embedded in different partitions inside the condensation tube wall, and each condensation structure includes a cooling water circulation system and an adjustable-angle deflector.

[0108] A calculation module 22, configured to calculate the condensation efficiency based on the annular temperature gradient distribution data and the phase change characteristic data.

[0109] A generation module 23, configured to input the condensation efficiency and the ambient temperature data into a pre-trained reinforcement learning parameter optimization model, and generate the flow rate adjustment parameter and the deflector deflection angle adjustment parameter for each partition in combination with the physical constraint conditions of the condensation structure.

[0110] An adjustment module 24, configured to convert the flow rate adjustment parameter and the deflector deflection angle adjustment parameter for each partition into corresponding electric control valve opening commands respectively, so as to synchronously adjust the opening of the control valve of the cooling water circulation system corresponding to different partitions and the angle of the adjustable-angle deflector.

[0111] Figure 2 The intelligent data monitoring and analysis system of a solar seawater desalination system described above can execute Figure 1 The intelligent data monitoring and analysis method of a solar seawater desalination system described in the embodiment shown. Its implementation principle and technical effects will not be elaborated. For the intelligent data monitoring and analysis system of a solar seawater desalination system in the above embodiment, the specific manners in which each module and unit perform operations have been described in detail in the embodiment related to the method, and will not be elaborated here.

[0112] In a possible design, Figure 2 The intelligent data monitoring and analysis system of a solar seawater desalination system in the embodiment shown can be implemented as a computing device, as Figure 3 shown. The computing device can include a storage component 31 and a processing component 32.

[0113] The storage component 31 stores one or more computer instructions, and one or more of the computer instructions are called and executed by the processing component 32.

[0114] The processing component 32 is configured to: during the condensation stage of the solar desalination system, obtain the annular temperature gradient distribution data of the condensation tube wall, the phase change characteristic data of the steam in the condensation tube during the condensation process, and the ambient temperature data, wherein corresponding condensation structures are embedded in different partitions inside the condensation tube wall, and each condensation structure includes a cooling water circulation system and an adjustable-angle deflector. Based on the annular temperature gradient distribution data and the phase change characteristic data, calculate the condensation efficiency. Input the condensation efficiency and the ambient temperature data into a pre-trained reinforcement learning parameter optimization model, and combine with the physical constraint conditions of the condensation structure to generate the flow rate adjustment parameters and the deflector deflection angle adjustment parameters for each partition. Convert the flow rate adjustment parameters and the deflector deflection angle adjustment parameters for each partition into corresponding electric control valve opening commands respectively, so as to synchronously adjust the opening degree of the control valve of the cooling water circulation system corresponding to different partitions and the angle of the adjustable-angle deflector.

[0115] Among them, the processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components, and is used to execute the above method.

[0116] The storage component 31 is configured to store various types of data to support the operation of the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as random access memory (RAM), static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0117] Of course, the computing device necessarily may also include other components, such as input / output interfaces, display components, communication components, etc.

[0118] The input / output interface provides an interface between the processing component and the peripheral interface module, and the peripheral interface module may be an output device, an input device, etc.

[0119] The communication component is configured to facilitate communication between the computing device and other devices in a wired or wireless manner, etc.

[0120] Among them, the computing device may be a physical device or an elastic computing host provided by a cloud computing platform, etc. At this time, the computing device may refer to a cloud server, and the above-mentioned processing component, storage component, etc. may be basic server resources leased or purchased from a cloud computing platform.

[0121] The embodiment of the present application also provides a computer storage medium storing a computer program, and when the computer program is executed by a computer, it can implement the above Figure 1 intelligent data monitoring and analysis method of a solar desalination system shown in the above embodiments.

[0122] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0123] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative labor.

[0124] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution essentially or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disc, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0125] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. An intelligent data monitoring and analysis method for a solar seawater desalination system, characterized in that, Including: During the condensation stage of the solar desalination system, obtain the annular temperature gradient distribution data of the condensation tube wall, the phase change characteristic data of the steam in the condensation tube during condensation, and the ambient temperature data. The corresponding condensation structures are embedded in different partitions inside the condensation tube wall, and each condensation structure includes a cooling water circulation system and an adjustable-angle deflector; Calculate the condensation efficiency based on the annular temperature gradient distribution data and the phase change characteristic data; Input the condensation efficiency and the ambient temperature data into a pre-trained reinforcement learning parameter optimization model, and combine the physical constraint conditions of the condensation structure to generate the flow rate adjustment parameters and the deflector deflection angle adjustment parameters for each partition; Convert the flow rate adjustment parameters and the deflector deflection angle adjustment parameters for each partition into corresponding electric control valve opening commands respectively, so as to synchronously adjust the opening of the control valve of the cooling water circulation system corresponding to different partitions and the angle of the adjustable-angle deflector.

2. The method according to claim 1, characterized in that, The step of inputting the condensation efficiency and the ambient temperature data into a pre-trained reinforcement learning parameter optimization model, and combining the physical constraint conditions of the condensation structure to generate the flow rate adjustment parameters and the deflector deflection angle adjustment parameters for each partition includes: Bind the condensation efficiency and the ambient temperature data in space-time according to the partition identification code, and input them into the feature extraction module of the reinforcement learning parameter optimization model to extract the dynamic feature sequence and the thermal inertia coefficient; Input the dynamic feature sequence into the dual-channel policy network of the reinforcement learning parameter optimization model. The first-channel policy network generates the initial flow rate parameters of the cooling water circulation system corresponding to each partition through a fully connected layer and a flow rate constraint layer, and the second-channel policy network generates the initial deflection angle parameters of the adjustable-angle deflector corresponding to each partition through a convolutional layer and an angle constraint layer. The flow rate constraint layer dynamically adjusts the constraint boundary based on the flow rate constraint strategy output by the reinforcement learning agent, and the angle constraint layer dynamically adjusts the constraint boundary based on the angle constraint strategy output by the reinforcement learning agent; Take the initial flow rate parameters and the initial deflection angle parameters as the state input of the reinforcement learning agent, receive the parameter correction actions output by the reinforcement learning agent, and perform parameter correction based on the physical constraint conditions to generate intermediate flow rate parameters and intermediate deflection angle parameters; Input the intermediate flow rate parameters and the intermediate deflection angle parameters into the multi-objective optimization module of the reinforcement learning parameter optimization model for parameter optimization; Cascade and fuse the optimized intermediate flow rate parameters, the optimized intermediate deflection angle parameters, and the historical optimal parameters at the partition level, and dynamically allocate the fusion weights according to the thermal inertia coefficients of the condensation structures in each partition to generate the flow rate adjustment parameters and the deflector deflection angle adjustment parameters for each partition.

3. The method according to claim 2, characterized in that, The step of inputting the dynamic feature sequence into the dual-channel policy network of the reinforcement learning parameter optimization model, where the first-channel policy network generates the initial flow rate parameters of the cooling water circulation system corresponding to each partition through a fully connected layer and a flow rate constraint layer, and the second-channel policy network generates the initial deflection angle parameters of the adjustable-angle deflector corresponding to each partition through a convolutional layer and an angle constraint layer, includes: Input each partition feature vector in the dynamic feature sequence into the fully connected layer, and map it to a traffic feature vector through dense connections between nodes; Input the traffic feature vector into the traffic constraint layer. The maximum pipe pressure threshold and the minimum flow velocity threshold of the cooling water circulation system are predefined in the traffic constraint layer. Perform element-wise truncation on the traffic feature vector through an inequality constraint function to generate the initial flow parameters of the cooling water circulation system corresponding to each partition; Input the two-dimensional spatial distribution features of each partition in the dynamic feature sequence into the convolutional layer, and perform sliding window convolution along the partition spatial dimension through multiple groups of convolutional kernels to extract the deflection angle correlation features of the flow guiding plates between adjacent partitions; Input the deflection angle correlation features of the flow guiding plates into the angle constraint layer. The upper and lower limits of the mechanical rotation angle of the adjustable angle flow guiding plate are predefined in the angle constraint layer. Map the deflection angle correlation features of the flow guiding plates to a preset angle interval through a piecewise linear mapping function to generate the initial deflection angle parameters of the adjustable angle flow guiding plates corresponding to each partition.

4. The method according to claim 1, characterized in that, Calculating the condensation efficiency based on the annular temperature gradient distribution data and the phase change characteristic data includes: Equally angularly divide the annular temperature gradient distribution data into a number of annular sub-regions in the radial direction of the annular region, and extract the temperature gradient vector data of each annular sub-region; Obtain the latent heat of phase change value and the upper limit value of the phase change temperature range of each annular sub-region from the phase change characteristic data, and calculate the heat response rate coefficient of each annular sub-region. The heat response rate coefficient is the ratio of the latent heat of phase change value of the corresponding annular sub-region to the upper limit value of the phase change temperature range; Multiply the temperature gradient vector data of each annular sub-region by the heat response rate coefficient of the corresponding annular sub-region to generate the condensation efficiency contribution value of each annular sub-region; Perform weighted summation on the condensation efficiency contribution values of each annular sub-region according to the area ratio of each annular sub-region in the annular region to generate the condensation efficiency.

5. The method according to claim 3, characterized in that, The step of inputting the traffic feature vector into the traffic constraint layer, where the maximum pipe pressure threshold and the minimum flow velocity threshold of the cooling water circulation system are predefined in the traffic constraint layer, and performing element-wise truncation on the traffic feature vector through an inequality constraint function to generate the initial flow parameters of the cooling water circulation system corresponding to each partition includes: Perform element-wise constraint processing on each element of the traffic feature vector. If the current element value is greater than the maximum allowable flow value corresponding to the maximum pipe pressure threshold of the corresponding partition, then modify the current element value to the maximum allowable flow value; If the current element value is less than the minimum allowable flow value corresponding to the minimum flow velocity threshold of the corresponding partition, then modify the current element value to the minimum allowable flow value; If the current element value is between the minimum allowable flow value and the maximum allowable flow value, then keep the current element value unchanged; Recombine the elements after element-wise constraint processing according to the partition mapping relationship to generate the initial flow parameters of the cooling water circulation system corresponding to each partition.

6. The method according to claim 2, characterized in that, Taking the initial flow rate parameter and the initial deflection angle parameter as the state input of the reinforcement learning agent, receiving the parameter correction action output by the reinforcement learning agent, and performing parameter correction based on the physical constraint conditions to generate an intermediate flow rate parameter and an intermediate deflection angle parameter, including: Numerically comparing the initial flow rate parameter with the upper limit value and the lower limit value of the constraint of the flow rate parameter respectively; Numerically comparing the initial deflection angle parameter with the upper limit angle and the lower limit angle of the constraint of the deflection angle parameter respectively; Determining the intermediate flow rate parameter according to the flow rate comparison result, and determining the intermediate deflection angle parameter according to the deflection angle comparison result.

7. The method according to claim 6, characterized in that, The determining the intermediate flow rate parameter according to the flow rate comparison result and determining the intermediate deflection angle parameter according to the deflection angle comparison result includes: If the flow rate comparison result indicates that the initial flow rate parameter is greater than the upper limit value of the constraint of the flow rate parameter, taking the first fixed offset as the intermediate flow rate parameter; if the flow rate comparison result indicates that the initial flow rate parameter is less than the lower limit value of the constraint of the flow rate parameter, taking the second fixed offset as the intermediate flow rate parameter; If the deflection angle comparison result indicates that the initial deflection angle parameter is greater than the upper limit angle of the constraint of the deflection angle parameter, generating an intermediate deflection angle parameter between the initial deflection angle parameter and the upper limit angle of the constraint of the deflection angle parameter through interpolation operation; if the initial deflection angle parameter is less than the lower limit angle of the constraint of the deflection angle parameter, generating an intermediate deflection angle parameter between the lower limit angle and the initial deflection angle parameter through interpolation operation.

8. An intelligent data monitoring and analysis system for a solar desalination system, characterized in that, Including: An acquisition module, configured to acquire the annular temperature gradient distribution data of the condensing tube wall, the phase change characteristic data of the steam in the condensing tube during the condensation process, and the ambient temperature data during the condensation stage of the solar desalination system, wherein corresponding condensing structures are embedded in different partitions inside the condensing tube wall, and each condensing structure includes a cooling water circulation system and an adjustable-angle deflector; A calculation module, configured to calculate the condensation efficiency based on the annular temperature gradient distribution data and the phase change characteristic data; A generation module, configured to input the condensation efficiency and the ambient temperature data into a pre-trained reinforcement learning parameter optimization model, and generate the flow rate adjustment parameter and the deflector deflection angle adjustment parameter of each partition in combination with the physical constraint conditions of the condensing structure; An adjustment module, configured to convert the flow rate adjustment parameter and the deflector deflection angle adjustment parameter of each partition into corresponding electric control valve opening commands respectively, so as to synchronously adjust the opening of the control valve of the cooling water circulation system corresponding to different partitions and the angle of the adjustable-angle deflector.

9. A computing device, characterized in that, Including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement an intelligent data monitoring and analysis method for a solar desalination system as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, Stored with a computer program, when the computer program is executed by a computer, it implements an intelligent data monitoring and analysis method for a solar desalination system as described in any one of claims 1 to 7.

Images