Direct spraying type waste heat recovery acidic water treatment method

Abstract

The invention relates to the technical field of water treatment, and discloses a direct spraying type waste heat recovery acidic water treatment method. The method comprises the following steps: arranging a pH value difference sensor group and a conductivity sensor at a joint of a buried pipe cold storage system and a direct-spraying type waste heat recovery device, and monitoring the whole process of acid water treatment in real time to obtain water quality change data of the acid water; performing filtering treatment and working condition matching analysis on the acid water quality change data and the system operation parameters to obtain a cold storage-acid water treatment cooperation mode; performing thermochemical coupling analysis through the hyperboloid mapping network to obtain a corrosion risk assessment result; and executing the heat exchange efficiency and pH value adjustment dual-constraint optimization model according to a corrosion risk assessment result, and outputting an anti-corrosion control strategy. According to the method, the complex nonlinear relation between the pH value change and the temperature gradient can be accurately captured, the accuracy of corrosion risk prediction is greatly improved, and damage to equipment caused by corrosion is effectively avoided.

Description

Technical Field

[0001] The present application relates to the field of water treatment technology, and in particular to a direct-spray waste heat recovery acidic water treatment method. Background Art

[0002] With the increasing demand for efficient energy utilization and environmental protection in industrial production processes, waste heat recovery technology has been widely used in various industries. In particular, in industries such as metallurgy, chemical industry, and electric power, the large amount of acidic wastewater generated by high-temperature production processes not only contains a large amount of waste heat, but is also highly corrosive. Traditional waste heat recovery systems are unable to effectively deal with the corrosion problem of acidic water. Although existing underground pipe cold storage technology can achieve cross-seasonal energy utilization, when combined with acidic water waste heat recovery, there is a contradiction between heat exchange efficiency and material corrosion. Currently, the industry's solutions mostly use the method of adding an intermediate heat exchange link to isolate the acidic medium, but this method significantly reduces energy utilization efficiency.

[0003] Traditional acidic water waste heat recovery systems primarily utilize fixed-parameter control strategies, lacking the ability to respond in real time to changes in the acidic water's pH value and temperature distribution. This makes it difficult for system operating parameters to adapt to changing operating conditions. Furthermore, existing monitoring systems have limited capabilities for assessing corrosion risks during acidic water treatment and are unable to achieve multi-objective collaborative optimization, resulting in suboptimal system operation. Furthermore, traditional control methods lack the ability to accumulate and learn data from long-term system operation, making it difficult to effectively address issues arising from equipment aging and performance degradation, limiting system stability and service life. Summary of the Invention

[0004] The present application provides a direct-spray waste heat recovery acidic water treatment method, which can accurately capture the complex nonlinear relationship between pH value changes and temperature gradients, greatly improve the accuracy of corrosion risk prediction, and effectively avoid damage to equipment caused by corrosion.

[0005] In a first aspect, the present application provides a direct-spray waste heat recovery acidic water treatment method, the direct-spray waste heat recovery acidic water treatment method comprising: A pH differential sensor group and a conductivity sensor are installed at the interface between the buried pipe cold storage system and the direct-spray waste heat recovery device to monitor the entire acidic water treatment process in real time and obtain data on changes in acidic water quality. Filtering and analyzing the acidic water quality change data and system operating parameters and operating condition matching to obtain a cold storage-acidic water treatment synergistic model; The cold storage-acid water treatment synergistic model and underground cold storage state data are input into a hyperbolic mapping network for thermochemical coupling analysis to obtain corrosion risk assessment results; A dual-constraint optimization model of heat exchange efficiency and pH value adjustment is executed according to the corrosion risk assessment result, and an anti-corrosion control strategy is output.

[0006] In the technical solution provided by the present application, by setting a high-precision monitoring device at the interface of the underground pipe cold storage system and the direct shower type waste heat recovery device, real-time monitoring of the entire treatment process is achieved, which significantly improves the monitoring accuracy and data reliability. The hyperbolic mapping network used for thermochemical coupling analysis can accurately capture the complex nonlinear relationship between pH value changes and temperature gradients, greatly improving the accuracy of corrosion risk prediction and effectively avoiding damage to equipment caused by corrosion. By establishing a dual-constraint optimization model of heat exchange efficiency and pH value adjustment, a balance is achieved between energy utilization efficiency, acidic water treatment effect and equipment corrosion protection. At the same time, the anti-corrosion control strategy based on risk assessment adopts coarse and fine adjustment two-stage pH value control, which effectively reduces pH value fluctuations and controls the corrosion rate within a safe range. The cold storage-acid water treatment synergistic mode constructed by the present invention gives full play to the energy and functional complementary advantages of the underground pipe cold storage system and the direct shower type waste heat recovery device, significantly reduces the system operating cost, reduces the amount of chemical reagents used, and improves the waste heat recovery rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0008] Figure 1 This is a schematic diagram of an embodiment of a direct-spray waste heat recovery acidic water treatment method in an embodiment of the present application. DETAILED DESCRIPTION

[0009] An embodiment of the present application provides a direct-spray waste heat recovery acidic water treatment method. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units that are not explicitly listed or inherent to these processes, methods, products or apparatus.

[0010] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiment of the present application, an embodiment of the direct-spray waste heat recovery acidic water treatment method includes: Step S101: Install a pH differential sensor group and a conductivity sensor at the interface between the buried pipe cold storage system and the direct shower waste heat recovery device to monitor the entire acidic water treatment process in real time and obtain acidic water quality change data; It is understood that the execution subject of this application can be a direct-spray waste heat recovery acidic water treatment system, or a terminal or a server, which is not limited here. The embodiment of this application is described by taking the server as the execution subject as an example.

[0011] Specifically, a pH differential sensor set and a conductivity sensor are installed at the interface between the underground pipe cold storage system and the direct-spray waste heat recovery device to provide real-time monitoring of the entire acidic water treatment process. A first pH sensor and a first conductivity sensor are installed between the outlet of the underground pipe cold storage system and the inlet of the direct-spray waste heat recovery device, establishing an inlet water parameter monitoring point. This provides inlet water characteristic data for the acidic water, reflecting basic water quality information before entering the waste heat recovery device. A second pH sensor and a second conductivity sensor are installed between the outlet of the direct-spray waste heat recovery device and the inlet of the acidic water collection tank, establishing an outlet water parameter monitoring point. This provides outlet water characteristic data for the acidic water, reflecting changes in water quality after treatment by the waste heat recovery device. A differential calculation is performed on the acidic water inlet and outlet characteristic data to obtain pH differentials (ΔpH) and conductivity differentials (ΔEC). This differential calculation reflects changes in water quality during treatment and identifies trends in acidic water during outlet water treatment. The differential data is correlated with data from the level and temperature sensors in the acidic water collection tank to form a multidimensional water quality parameter matrix. This matrix includes changes in pH and conductivity, as well as other factors influencing water quality, such as level and temperature. This matrix is ​​arranged in a time series format and processed using a sliding window algorithm to extract dynamic trends within the time series, eliminating the influence of short-term fluctuations. This yields more stable and regular water quality data, reflecting immediate changes in acidic water quality and revealing long-term trends during the treatment process, enabling continuous monitoring and regulation of water quality. The dynamic water quality trend data is then compared against the pre-set acidic water treatment standard range, and outliers are flagged. This comparative analysis process quickly identifies water quality changes outside the normal range, allowing for timely action to address any issues.

[0012] Step S102: filtering and analyzing the acidic water quality change data and system operating parameters, and obtaining a cold storage-acidic water treatment synergistic mode; Specifically, the pH and conductivity data from the acidic water quality change data were reorganized into time series to generate an initial water quality feature sequence, reflecting the fluctuations in acidic water quality during the treatment process. Furthermore, to better understand the actual system operation, real-time data preprocessing was performed on the temperature and pressure sensors in the underground pipe cold storage system to generate a system operation feature sequence. The initial water quality feature sequence was merged with the system operation feature sequence to generate an original coupled data matrix. This data matrix combines information on acidic water quality changes with the system's operating status, providing a more comprehensive picture of system operation. Because the data contains a certain amount of noise, an adaptive Kalman filter algorithm was used to dynamically suppress the noise in the original coupled data matrix to reduce unnecessary interference. This process produces a low-noise data stream, ensuring data accuracy and reliability. Based on the temporal relationship between pH values ​​in the low-noise data stream and changes in system temperature and pressure, a multivariable transfer function was constructed to develop an acid-base-temperature-pressure response model, describing the relationship between acidic water quality changes and changes in temperature and pressure in the underground pipe cold storage system. Parameter identification and sensitivity analysis of the acid-base-temperature-pressure response model were performed to extract target influencing factors and their quantitative relationships. This helped identify key factors influencing water quality changes and system operation, and quantitatively analyzed the changing patterns of these factors to generate acid-base balance correlation data. This correlation data helped understand the interactions between different factors. Based on this acid-base balance correlation data, the refrigerant-acid water heat transfer index and pH adjustment coefficient were calculated, and an operating condition matching analysis was performed. This analysis optimized the coordinated model of cold storage and acid water treatment based on water quality changes and system operating conditions under different operating conditions, ensuring optimal system operation under different operating conditions.

[0013] In this embodiment, key parameters are extracted from the acid-base balance correlation data, including temperature difference, pH change rate, and conductivity change value. These three parameters serve as important indicators reflecting the dynamics of water quality and thermal conductivity efficiency. By combining them with the heat exchange theoretical model, a heat exchange efficiency function based on water quality and thermodynamic coupling is constructed. This function comprehensively reflects the energy exchange efficiency between acidic water and the refrigerant in the direct-spray waste heat recovery device, calculates the refrigerant-acidic water heat exchange index, and provides a quantitative basis for evaluating the system's heat recovery performance. In order to achieve stable water quality control, based on the deviation information between the current pH value recorded in the acid-base balance correlation data and the pH target value set by the system, an adaptive adjustment algorithm is executed through dynamic tracking and error feedback mechanism to calculate the pH adjustment coefficient. This coefficient represents the response intensity and adjustment rate of the system's neutralization and regulation ability to acidic water, and is a key control parameter for maintaining the chemical stability of the system. The obtained refrigerant-acid water heat transfer index and pH adjustment coefficient are combined with the operating state parameters of the underground pipe cold storage system (such as current cooling load, flow rate, and pressure) to form an operating condition matching matrix. This matrix reflects the comprehensive performance of the system under various operating conditions. To identify and categorize system operating modes, a fuzzy clustering analysis method is applied to this operating condition matching matrix. Each operating state is fuzzily classified based on similarity, forming a system operating condition classification. Based on the classified operating states, a state transition probability matrix is ​​constructed to quantify the probability of the system transitioning from one operating mode to another. A Markov decision process is then applied to optimize the policy for all possible states. By finding the behavior with the highest long-term benefit in each state, the optimal policy function is derived, which is then used to form a mode switching decision table. This mode switching decision table is then cascade-matched with the aforementioned refrigerant-acid water heat transfer index and pH adjustment coefficient. Calculation and fusion analysis are then performed to determine the optimal coordinated operating strategy under different system states and water quality conditions, outputting the final cold storage-acid water treatment coordinated mode. This mode can achieve a dynamic balance between waste heat recovery efficiency and water quality safety control, allowing the system to maintain efficient and stable operation under complex and changing working conditions.

[0014] Step S103: Input the cold storage-acid water treatment collaborative model and the underground cold storage state data into a hyperbolic mapping network for thermochemical coupling analysis to obtain a corrosion risk assessment result; Specifically, core parameters representing the system's operational characteristics are extracted from the constructed cold storage-acid water treatment synergistic model. These parameters include a condition type parameter representing the operational category, a heat transfer performance index measuring the system's energy efficiency, and a pH adjustment quantification reflecting chemical regulation behavior. Together, these parameters constitute the synergistic model's characteristic data, describing the system's dynamic state at both the thermal and chemical levels. Furthermore, to comprehensively reflect underground environmental conditions, real-time soil temperature distribution and local heat flux data are collected from temperature sensors deployed in the buried area of ​​the buried pipe-in-pipe heat exchanger. This data constitutes the underground cold storage state data, characterizing the potential impact of the underground thermophysical background on heat transfer processes and chemical reactions. These synergistic model characteristic data and underground cold storage state data are fed into a hyperbolic mapping network. Its hyperbolic embedding layer achieves a nonlinear mapping from Euclidean space features to hyperbolic space, compressing the originally high-dimensional and complex physical and chemical variables into a more expressive nonlinear dimensionality reduction representation. This mapping enhances the geometric representation of the data while preserving the structural relationships between multiple variables, providing a more compact feature representation for subsequent analysis. The nonlinear dimensionality reduction representation is input to the network's thermochemical coupling layer, where a hyperbolic convolution operation is performed to extract the interaction between the dynamic changes in acidic water pH and the underground temperature gradient. This layer constructs key physical-chemical coupling factors in the corrosion process, ultimately outputting corrosion driving factors. These factors are the fundamental source of the system's potential corrosion risk. The corrosion driving factors are then fed into the risk quantification layer of the hyperbolic mapping network. Through multidimensional feature fusion and electrochemical model parameter calculation, the corrosion reaction activity at the material-environment interface is quantitatively calculated, resulting in a spatially discretized corrosion rate distribution. This corrosion rate distribution reflects the differences in corrosion intensity at different locations and serves as the basis for time-based extrapolation. To predict corrosion evolution trends over time, this corrosion rate distribution is extrapolated over time and compared with the corrosion resistance threshold of the material being used. A distance metric function in hyperbolic space is used to assess whether the corrosion rate approaches or exceeds the material's safety limit. The corresponding corrosion risk level is then calculated and ultimately output as a corrosion risk assessment result for the system.

[0015] Step S104: Execute a dual-constraint optimization model of heat exchange efficiency and pH value adjustment according to the corrosion risk assessment result, and output an anti-corrosion control strategy.

[0016] Specifically, the risk levels in the corrosion risk assessment results are standardized and mapped to corresponding corrosion protection constraints. These constraints include physical and chemical boundary constraints such as upper limits on corrosion rates, critical pH ranges, and maximum allowable stress values ​​for key system components. This conversion process transforms the originally qualitative or hierarchical corrosion risk information into a computable optimization framework, serving as the foundation for subsequent modeling. When developing the optimization model, the corrosion protection constraints serve as hard boundaries, and a composite objective function is constructed that incorporates multiple system operational objectives. The energy efficiency objective function evaluates the heat transfer efficiency per unit of system energy consumption; the acidic water treatment efficiency objective function measures the reaction completion and water quality stability of the acidic water neutralization process; and the system operating cost objective function considers the impact of factors such as reagent dosage, equipment load, and maintenance frequency on overall costs. These three objective functions are integrated into an optimization target model to ensure a balance between operational efficiency and economic efficiency while ensuring system corrosion safety. A control parameter space is defined, encompassing adjustable operational variables such as coolant flow rate, pH adjuster dosage rate, and the setpoint for the inlet and outlet temperature difference of the heat exchanger. Based on a defined control parameter space and a composite objective model, a non-dominated sorting genetic algorithm (NSA) is introduced for multi-objective optimization. This algorithm continuously generates candidate solutions through a simulated evolutionary process, selecting solutions that are mutually non-dominated across the three objective dimensions to form the Pareto optimal solution set—the set of optimal control combinations under all trade-offs. Because each solution in the Pareto solution set performs differently on different objectives, a fuzzy membership function is used to evaluate the satisfaction of each solution. This function assigns a satisfaction score between [0 and 1] based on the degree of proximity between the actual value of each objective function and the expected value. The control parameter combination with the highest overall satisfaction is selected through weighting or normalization as the optimal control solution for system operation. After obtaining the optimal control parameter combination, the pH adjustment process is refined into two stages: a coarse adjustment stage and a fine adjustment stage. In the coarse adjustment stage, a larger adjustment step size is used based on the deviation amplitude to quickly adjust the pH value to close to the target range. In the fine adjustment stage, a smaller pH adjustment coefficient is used to stabilize the pH value with more precise adjustments to prevent overshoot or fluctuations in the system. The entire regulation process is based on real-time feedback from the system, dynamically adjusting the regulation range and cycle. Ultimately, combining heat exchange efficiency control requirements with pH value fine-tuning logic, a complete anti-corrosion control strategy is automatically generated.

[0017] Key operational information, including real-time temperature, pressure, flow rate, and acidity data, is collected from target monitoring points within the underground pipe cold storage system. These parameters constitute the multivariate inputs to the current system operating state. Preprocessing operations, including outlier removal, data smoothing, and missing value interpolation, are performed on the raw real-time state parameters. Normalization is then used to convert all parameters to the same interval, resulting in a numerically stable and scaled matrix for the system's current operating state. Based on this matrix, combined with the optimal control parameter combination from the previously constructed anti-corrosion control strategy, the system's current operating state matrix and the control strategy parameters are combined to form an upper-level proxy input vector. This vector reflects the system's physical operating state and incorporates strategic control information, enabling it to describe the global control logic. To prevent redundant features from interfering with subsequent modeling, a feature selection algorithm is used to filter this proxy input vector, extracting key variables highly correlated with control behavior and removing noisy and weakly correlated features. This results in proxy-level decision data with strong predictive power and control directionality. The upper-level decision function performs a global policy analysis on the agent-level decision data. This upper-level decision function, constructed based on a reinforcement learning value function, support vector regressor, or adaptive fuzzy logic model, aims to identify the optimal adjustment direction while maintaining system stability and efficiency. By calculating the impact weights of each input feature on system performance indicators (such as heat exchange efficiency, corrosion risk, and pH stability), the system's current adjustment direction is output, determining whether to enhance pH neutralization, accelerate heat release, or reduce system load. Based on the system adjustment direction output by the upper-level policy analysis function, the execution layer further refines the optimization objectives and identifies the specific control variables to be optimized in this round of adjustments, such as refrigerant circulation flow, acid-base regulator concentration, valve opening, or heat exchanger inlet and outlet temperature differential control values. This process ultimately outputs precise direct-injection waste heat recovery control parameters, enabling adaptive adjustment of multiple objectives under dynamic and complex operating conditions.

[0018] In the embodiment of the present application, by setting a high-precision monitoring device at the interface of the underground pipe cold storage system and the direct shower type waste heat recovery device, real-time monitoring of the entire treatment process is achieved, which significantly improves the monitoring accuracy and data reliability. The hyperbolic mapping network used for thermochemical coupling analysis can accurately capture the complex nonlinear relationship between pH value changes and temperature gradients, greatly improving the accuracy of corrosion risk prediction and effectively avoiding damage to equipment caused by corrosion. By establishing a dual-constraint optimization model of heat exchange efficiency and pH value adjustment, a balance is achieved between energy utilization efficiency, acidic water treatment effect and equipment corrosion protection. At the same time, the anti-corrosion control strategy based on risk assessment adopts coarse and fine adjustment two-stage pH value control, which effectively reduces pH value fluctuations and controls the corrosion rate within a safe range. The cold storage-acid water treatment synergistic mode constructed by the present invention gives full play to the energy and functional complementary advantages of the underground pipe cold storage system and the direct shower type waste heat recovery device, significantly reduces the system operating cost, reduces the amount of chemical reagents used, and improves the waste heat recovery rate.

[0019] In a specific embodiment, the process of executing step S101 may specifically include the following steps: A first pH sensor and a first conductivity sensor are installed between the water outlet of the buried pipe cold storage system and the water inlet of the direct-spray waste heat recovery device to establish an inlet parameter monitoring point and obtain acidic water inlet characteristic data; A second pH sensor and a second conductivity sensor are installed between the water outlet of the direct-spray waste heat recovery device and the inlet of the acidic water collection tank to establish an outlet water parameter monitoring point and obtain acidic water outlet characteristic data; Perform differential calculation on the acidic water inlet characteristic data and the acidic water outlet characteristic data to obtain the pH value difference ΔpH and the conductivity difference ΔEC value; The pH difference ΔpH and conductivity difference ΔEC values ​​were correlated with the data from the liquid level sensor and temperature sensor in the acidic water collection tank to obtain a multidimensional water quality parameter matrix. The multidimensional water quality parameter matrix is ​​arranged in time series and processed using a sliding window algorithm to obtain dynamic water quality change trend data; The dynamic water quality change trend data is compared and analyzed with the preset acidic water treatment standard range, and abnormal points are marked to obtain the acidic water quality change data.

[0020] Specifically, a first pH sensor and a first conductivity sensor are installed between the outlet of the underground pipe cold storage system and the inlet of the direct-spray waste heat recovery device to establish an inlet water parameter monitoring point. This monitoring point captures the water quality of the acidic water discharged from the underground pipe system in real time before it enters the direct-spray device. The pH sensor measures the acidity and alkalinity of the inlet water, while the conductivity sensor senses the ion concentration level in the water. Together, the two sensors form a basic characteristic dataset for the inlet water quality. Simultaneously, a second pH sensor and a second conductivity sensor are installed between the outlet of the direct-spray waste heat recovery device and the inlet of the acidic water collection tank to establish an outlet water parameter monitoring point, enabling continuous monitoring of the pH and conductivity of the treated acidic water. This arrangement ensures that any changes in water quality during its passage through the direct-spray device are captured. This sensor setup provides a complete inlet-outlet water quality data stream, which is then used to differentially calculate the inlet and outlet water characteristic data to obtain the pH difference ΔpH and the conductivity difference ΔEC. These two differential values ​​reflect the effectiveness of the direct-injection waste heat recovery system in regulating and purifying acidic water during treatment. ΔpH characterizes the degree of change in hydrogen ion concentration in the water, reflecting its pH-regulating capacity; ΔEC quantifies ion migration behavior, revealing the removal or changing trends of salts, metal ions, or other conductive factors in the system. Based on these ΔpH and ΔEC, liquid level and temperature data are simultaneously acquired from the acidic water collection tank, where level and temperature sensors are installed. Liquid level data is used to infer the balance between inflow and outflow rates, indirectly assessing treatment load and storage capacity. Temperature data reveals the heat exchange behavior between water quality and the environment, and is correlated to changes in waste heat recovery efficiency. The four heterogeneous data types—ΔpH, ΔEC, liquid level, and temperature—are fused through time-series synchronization, data alignment, and dimensionality unification to construct a multidimensional water quality parameter matrix. This matrix records various characteristics of the current system water quality during each sampling period, forming a dynamic, time-series data framework. To reveal the changing trends of acidic water quality during operation, the multidimensional water quality parameter matrix was arranged in a time series to construct a characteristic sequence reflecting the system's operating history. To mitigate the influence of short-term disturbances or outliers on trend analysis, a sliding window algorithm was introduced to process the data based on the sequence. The sliding window algorithm's design considers the impact of window length and step size on the sensitivity and stability of trend extraction results. It extracts local data segments over continuous time slices and aggregates, smoothes, or performs coefficient of variation analysis on the data within these segments to form a set of time-sliding state curves. This method eliminates the risk of misjudgment caused by short-term, high-frequency disturbances and effectively captures medium- and long-term trend changes, helping to identify potential water quality hazards such as accelerated corrosion, unstable pH fluctuations, and cumulative conductivity increases. The dynamic water quality trend data generated by the sliding window algorithm was then compared against the pre-set acidic water treatment standard range.Standard ranges typically include specified pH limits, conductivity thresholds, temperature limits, and liquid level warning lines, used to determine whether the system is operating under safe, stable, and controllable conditions. The comparative analysis process uses threshold functions and tolerance ranges to dynamically identify errors and flag any data segments that fail to meet the pre-set standards as abnormal. Abnormality flags range from minor fluctuation alerts to severe deviation warnings and trend deterioration warnings. Event labels are generated based on the degree of deviation. These abnormal points, along with all trend data, are defined as acidic water quality change data.

[0021] In a specific embodiment, the process of executing step S102 may specifically include the following steps: The pH value and conductivity in the acidic water quality change data are reorganized into time series to obtain the initial water quality feature sequence. The temperature and pressure sensors in the buried pipe cold storage system are preprocessed in real time to obtain the system operation feature sequence. The initial water quality characteristic sequence and the system operation characteristic sequence are merged to obtain the original coupling data matrix; Adaptive Kalman filter algorithm is used to dynamically suppress the noise of the original coupled data matrix to obtain a low-noise data stream; A multivariable transfer function is constructed based on the temporal relationship between pH value and system temperature and pressure changes in the low-noise data stream to obtain an acid-base-temperature-pressure response model; Perform parameter identification and sensitivity analysis on the acid-base-temperature-pressure response model, extract target influencing factors and their quantitative relationships, and obtain acid-base balance correlation data; The refrigerant-acidic water heat exchange index and pH adjustment coefficient were calculated based on the acid-base balance correlation data, and the operating condition matching analysis was performed to obtain the cold storage-acidic water treatment synergistic mode.

[0022] Specifically, the time series of pH and conductivity, the core variables in the acidic water quality data, were reorganized. By reorganizing these two key water quality parameters in chronological order, an initial water quality feature sequence reflecting the water quality trend was constructed. This sequence preserves the trajectory of each variable over a continuous time slice, and normalization ensures that each dimension has the same metric scale in subsequent analysis. Furthermore, to comprehensively evaluate the driving effect of system operating conditions on water quality responses, real-time operating data was collected from temperature and pressure sensors deployed in the underground pipe cold storage system. After acquisition, this raw data underwent a series of preprocessing operations, including denoising, outlier removal, trend smoothing, normalization, and missing data interpolation, to generate a system operating feature sequence. This sequence, in the form of a continuous, uninterrupted data stream, depicts the system's operating state over a given time period, particularly its dynamic behavior with respect to temperature gradients and pressure fluctuations. This operating feature sequence complements the initial water quality feature sequence, with the former representing physical states and the latter reflecting chemical properties. The fusion of the two forms the basis for subsequent thermochemical coupling analysis. After synchronizing and structurally matching the initial water quality characteristic sequence with the system operation characteristic sequence in the temporal dimension, an original coupled data matrix is ​​constructed. Each row of this matrix corresponds to a time slice, and each column corresponds to a variable, covering the state values ​​of indicators such as pH, conductivity, system temperature, and pressure at each moment. Dynamic noise suppression is performed on the original coupled data matrix using an adaptive Kalman filter algorithm. This algorithm establishes a state prediction model and an observation model and automatically adjusts the noise covariance matrix based on prediction errors to achieve an optimal estimate of the system's true state. The filtering process preserves the dynamic trends of the data while suppressing unstructured noise caused by sudden changes, electromagnetic interference, or sensor drift, resulting in a low-noise data stream. Based on this low-noise data stream, the dynamic response relationship between pH changes and system temperature and pressure is analyzed, and a multivariable transfer function model is constructed. A transfer function model is a mathematical tool used to characterize the dynamic causal relationship between input and output variables. In this model, system temperature and pressure are used as input variables, and pH is used as the output variable. Model parameters are solved using methods such as least squares estimation, recursive least squares, or frequency domain fitting. This acid-base-temperature-pressure response model reveals the path of thermal drive's influence on acid-base regulation behavior and quantifies the transmission delay and gain effects of temperature gradients and pressure fluctuations on pH stability. By establishing this model, a causal mapping from physical states to chemical changes is achieved. Parameter identification and sensitivity analysis were performed on the acid-base-temperature-pressure response model. The parameter identification process used the minimum error inversion method to obtain the quantitative influence coefficient of each control parameter on the pH output. Sensitivity analysis, using methods such as partial derivatives, perturbation response methods, or Sobol global sensitivity analysis, identified the input factors most sensitive to pH regulation during system operation.The results of this process are summarized as acid-base balance correlation data, which includes the weighted contributions of each input parameter, the nonlinear interactions between variables, and the control range within the stable operating range. These data reflect the coupling characteristics between the physical system's operating state and its acidic water treatment capacity. Based on this acid-base balance correlation data, operating condition adaptability analysis and performance quantification are performed. The refrigerant-acidic water heat transfer index is calculated. This metric reflects the system's thermal conversion efficiency based on the heat transfer capacity between the refrigerant and acidic water per unit temperature difference, taking into account factors such as temperature gradient, pH response speed, and fluid thermal conductivity. Furthermore, the pH adjustment coefficient is calculated based on the system's current dynamic pH adjustment response range and adjustment speed. This coefficient reflects the adjustment energy consumption and response time required to restore the system to the target pH range under a unit disturbance, indirectly quantifying the system's chemical stability control capability. The heat transfer index and pH adjustment coefficient are input into the operating condition matching analysis module. By scoring the degree of matching with different historical operating conditions, the system's most similar operating mode is identified. The optimal coordination strategy is selected based on pre-set rules, ultimately resulting in a coordinated cold storage-acidic water treatment mode.

[0023] In a specific embodiment, the step of calculating the refrigerant-acidic water heat exchange index and pH adjustment coefficient based on the acid-base balance correlation data, and performing a working condition matching analysis to obtain the cold storage-acidic water treatment synergistic mode may specifically include the following steps: The temperature difference, pH change rate and conductivity change value are extracted from the acid-base balance correlation data, and the heat transfer efficiency is calculated based on the temperature difference, pH change rate and conductivity change value to obtain the refrigerant-acid water heat transfer index; According to the deviation between the current pH value and the target pH value in the acid-base balance associated data, adaptive adjustment is performed to obtain a pH adjustment coefficient; The refrigerant-acid water heat transfer index, pH value adjustment coefficient and system operating state parameters are combined into an operating condition matching matrix, and fuzzy cluster analysis is performed on the operating condition matching matrix to obtain the system operating condition classification; Based on the classification of system operating conditions, a state transition probability matrix is ​​constructed, and the optimal policy function is calculated using the Markov decision process to obtain the mode switching decision table; The mode switching decision table is matched with the refrigerant-acid water heat exchange index and pH value adjustment coefficient to obtain the cold storage-acid water treatment synergistic mode.

[0024] Specifically, core variables reflecting the system's heat and mass exchange state and chemical reaction efficiency are extracted from acid-base equilibrium correlation data. The temperature difference represents the thermodynamic driving force in the heat exchange process between the refrigerant and acidic water and is an important basis for evaluating the heat exchange rate. The pH change rate characterizes the dynamic adjustment of the pH value per unit time, revealing the intensity of the chemical response during the neutralization process in acidic water. The conductivity change reflects the trend of ion concentration in the water and is related to the migration and reaction state of conductive factors such as salt and metal ions in the water. These three variables are normalized and time-synchronized before serving as inputs for a comprehensive evaluation of heat exchange efficiency. The heat exchange efficiency calculation process constructs a multi-factor function based on these three variables. The temperature difference reflects the heat flow intensity, the pH change rate represents the driving rate of the chemical reaction, and the conductivity change indirectly characterizes the changes in the physical properties of the reaction products. By integrating these three variables, a heat exchange efficiency index function is established. This function adopts a weighted linear model or nonlinear regression model. After parameter identification and weight optimization, it outputs a unified quantitative indicator, namely the refrigerant-acidic water heat exchange index. This index, expressed in dimensionless form and comparable across different operating conditions, reflects both the heat conversion efficiency and, indirectly, the mass transfer capacity during the reaction. Simultaneously, an adaptive adjustment strategy is implemented based on the deviation between the current pH value recorded in the acid-base balance data and the set pH target value. This strategy utilizes a feedback controller to sense the current deviation in real time and determine whether it exceeds a preset tolerance. If the deviation is significant, the system increases the adjustment amplitude to quickly approach the target pH value. If the deviation is small, the system enters a fine-tuning state, reducing the fluctuation amplitude to prevent overshoot. A variable, the pH adjustment coefficient, is calculated to quantify the adjustment rate and intensity. The refrigerant-acid water heat transfer index, the pH adjustment coefficient, and system operating state parameters are combined to form a condition matching matrix. This matrix organizes data in time slices, with each row corresponding to the current operating state, encompassing multiple dimensions such as heat transfer efficiency, chemical adjustment capacity, temperature, pressure, flow rate, and conductivity. Each column represents a condition variable dimension. Each element of the matrix evaluates the degree of match between the system operating state and the target control range. In order to effectively classify the system's operating conditions, a fuzzy clustering analysis method is applied to the operating condition matching matrix. Fuzzy clustering allows each state to belong to multiple categories with different membership degrees, and can characterize the transition state behavior of the system when operating in the fuzzy boundary region. Using clustering techniques such as the fuzzy C-means algorithm or the Gaussian mixture model, the operating conditions are divided into several categories, such as high heat transfer and high stability, low heat transfer and high deviation, medium heat transfer and medium fluctuation, etc., to form a classification map of the system's operating conditions. Based on the operating condition classification results obtained by fuzzy clustering, a state transition probability matrix is ​​constructed. This matrix uses the operating condition category as the state node and describes the probability of the system transitioning from one operating condition to another between different time slices.The state transition matrix is ​​obtained by calculating the state transition frequency using a sliding window and normalizing it. It is essentially a Markov state transition model with a memoryless nature, enabling prediction of system operating trends over time. Based on this, a Markov decision process model is introduced. By defining the state space, action set, reward function, and state transition probability, the optimal action strategy for each operating state is determined. A policy iteration or value iteration algorithm is used to calculate the optimal policy function. This function outputs the control parameter adjustment direction and target combination to be executed under the current operating condition classification, and this function forms a mode switching decision table based on this function. This decision table uses the operating condition classification as input and outputs heat exchange parameter adjustment, pH adjustment priority, and energy load adjustment strategy, completing a logical loop from state identification to parameter decision making. This mode switching decision table is then combined with the calculated refrigerant-acid water heat exchange index and pH adjustment coefficient for parameter cascade and condition matching. The matching process adopts heuristic search or fuzzy rule reasoning system to select the optimal strategy path and dynamically adjust the control parameters according to the comprehensive score of heat exchange intensity and regulation capacity under the current operating conditions, and finally output the cold storage-acid water treatment collaborative mode that adapts to the current thermochemical state.

[0025] In a specific embodiment, the process of executing step S103 may specifically include the following steps: According to the cold storage-acid water treatment synergistic mode, the operating condition type parameters, heat transfer performance indicators and pH adjustment quantitative values ​​are extracted as the synergistic mode characteristic data; The soil temperature distribution and heat flux density data are collected from the temperature sensor at the buried location of the buried pipe-in-pipe heat exchanger as underground cold storage status data; The collaborative mode feature data and underground cold storage state data are input into the hyperbolic embedding layer of the hyperbolic mapping network. The Euclidean space features are mapped to the hyperbolic space through the hyperbolic embedding layer to obtain a nonlinear dimensionality reduction representation. The nonlinear dimensionality reduction representation is input into the thermochemical coupling layer of the hyperbolic mapping network. The thermochemical coupling layer performs a hyperbolic convolution operation to extract the interaction characteristics between pH value changes and temperature gradients to obtain the corrosion driving factors. The corrosion driving factors are input into the risk quantification layer of the hyperbolic mapping network. The electrochemical reaction activity of the material-environment interface is calculated through the risk quantification layer to obtain the spatially discretized corrosion rate distribution. The corrosion rate distribution is extrapolated in time series and compared with the material corrosion resistance threshold. The risk level is calculated using the hyperbolic distance metric function to obtain the corrosion risk assessment result.

[0026] Specifically, in constructing the corrosion risk assessment pathway, three key characteristic parameters were extracted from the established cold storage-acid water treatment synergistic model. The operating condition parameter primarily characterizes the control strategy or load mode of the current system operation, such as high-load, partial-load, and low-load conditions, and is used to determine the boundary conditions and control logic of the system operation. The heat transfer performance index quantifies the heat exchange capacity between the acid water and the refrigerant, reflecting the system's heat transfer efficiency per unit time. The pH adjustment quantification value, a numerical parameter reflecting the chemical control accuracy, reflects the intensity and frequency of the system's regulation in response to acid-base perturbations. These three parameters were normalized and time-aligned to form a synergistic model feature dataset, which captures the state of the system's thermochemical coupling behavior during each sampling period. Furthermore, given that the efficiency of the heat exchange process and the occurrence of corrosion reactions are highly dependent on the geothermal environment in which the heat exchanger is located, environmental parameters in the buried area of ​​the buried pipe-in-pipe heat exchanger, particularly soil temperature distribution and heat flux density, were collected in real time. This underground cold storage status data is acquired in real time via a spatial sensing network consisting of a temperature sensor array and a heat flux meter. It reflects the thermal conductivity of the underground medium and the heat capacity characteristics of the surface heat exchange process. Interpolation and spatial reconstruction of this data create a continuous three-dimensional temperature field and heat flux distribution map, providing the physical context for subsequent identification of thermally driven processes. The collaborative pattern feature data and underground cold storage status data are input into the hyperbolic embedding layer of the hyperbolic mapping network. This embedding layer maps the high-dimensional, unstructured Euclidean space feature data into a more expressive hyperbolic space. Due to its superior ability to model hierarchical relationships and capture nonlinear deformations, hyperbolic space offers significant advantages in addressing multivariate, nonlinear, and weakly coupled problems in systems. In the embedding layer, feature projection functions based on hyperbolic geometric structures, such as hyperbolic bilinear transformations, hyperbolic activation functions, and negative curvature mapping mechanisms, are constructed to achieve nonlinear dimensionality reduction of the input data while preserving the original structural relationships. The resulting output is a nonlinear dimensionality-reduced representation. The nonlinear dimensionality reduction representation is input into the core computing module of the hyperbolic mapping network: the thermochemical coupling layer. In this layer, the model performs a hyperbolic convolution operation and extracts interactive features of the pH value change trend and the underground temperature gradient change by constructing a multidimensional convolution kernel in the hyperbolic geometric tensor space. The model can identify the dynamic relationship between the time sensitivity of pH changes and the spatial continuity of temperature gradient changes, quantify the degree of coupling between the two, and extract a set of representative corrosion driving factors. The essence of the corrosion driving factor is a set of variables with spatial topological structure and physical quantity interpretation, which describes the core influencing source of metal corrosion and material degradation under the current operating state of the system. The corrosion driving factor is input into the third functional layer of the network: the risk quantification layer. At this stage, the model realizes the quantitative prediction of the possibility of corrosion by constructing the electrochemical reaction activity function and the material interface reaction mechanism model.This layer considers multiple variables, including metal surface activity, acid-base neutralization reaction rates, and conductivity coupling behavior, and combines them with reaction kinetic parameters to output an electrochemical reactivity index at the material-environment interface. Based on this index and the discretized spatial coordinates of the underground heat exchanger surface, a spatially discretized corrosion rate distribution map is reconstructed for the surfaces of key components throughout the system. This distribution is used to identify high-risk corrosion areas and assess the temporal and spatial cumulative effects of different operating conditions on corrosion in the system. Time series extrapolation of the corrosion rate distribution is performed. By constructing a sliding window-based time series prediction model, interpolation and fitting are performed based on the corrosion rate trends of several recent time slices to estimate the corrosion rate evolution over several future time steps. The extrapolated corrosion rate curve is compared with the material's preset corrosion resistance threshold. In areas where the local rate is about to exceed the safety upper limit, the model triggers a risk level calculation mechanism. This mechanism relies on a hyperbolic distance metric function, which measures the geometric distance between the current feature representation and the risk boundary in hyperbolic space. Smaller distance values ​​indicate higher corrosion risk; values ​​approaching zero indicate critical corrosion risk. Using the distance threshold as a reference and combining it with risk classification criteria, the system outputs a corrosion risk assessment result, including risk level identification, location of high-risk areas, and recommended control strategies.

[0027] In a specific embodiment, the process of executing step S104 may specifically include the following steps: Convert the risk level in the corrosion risk assessment results into corrosion protection constraints; Based on the corrosion protection constraints, an optimization target model is constructed, which includes the energy utilization efficiency objective function, the acid water treatment efficiency objective function and the system operation cost objective function. Define the control parameter space according to the optimization target model, and perform non-dominated sorting genetic solution based on the control parameter space and the optimization target model to obtain the Pareto optimal solution set; Each solution in the Pareto optimal solution set is evaluated using a fuzzy membership function to obtain a satisfaction score, and the optimal control parameter combination is selected based on the satisfaction score; Based on the optimal control parameter combination, the pH adjustment process is divided into two stages: coarse adjustment and fine adjustment. Different adjustment ranges are designed in combination with the pH adjustment coefficient, and an anti-corrosion control strategy is generated.

[0028] Specifically, the risk levels contained in the corrosion risk assessment results are given in the form of graded labels, distance thresholds, or rate intervals. These labels are converted into corrosion protection constraints and mapped by setting upper corrosion rate values, pH adjustment rate ranges, maximum allowable temperature gradients, or heat exchanger surface stress limits. On this basis, each level of corrosion risk is quantified into specific numerical or logical constraints, which serve as boundary conditions in the subsequent optimization process. Within the constraint domain of the corrosion protection constraints, a multi-objective optimization model is constructed to guide the system in seeking the optimal balance between multiple performance indicators. The model employs three objective functions. The first is an energy efficiency objective, which aims to minimize energy consumed per unit cooling load or heat exchange rate. This objective uses metrics such as heat exchange efficiency, coefficient of performance (COP), or energy consumption per unit cooling capacity. The second is an acidic water treatment efficiency objective, which emphasizes maintaining the acidic water pH within a target range during system operation while ensuring the appropriate dosage of chemical neutralizers and enhancing the reaction rate. The third is a system operating cost objective, which comprehensively considers the total cost of electricity, water, and chemical consumption, as well as the equipment operating load. This objective plays a decisive role in the system's economic viability. With these three objectives working together, the constructed optimization model focuses on maximizing system efficiency while maintaining a dynamic balance between corrosion control and economic operation. Based on this optimization objective model, a control parameter space is defined, encompassing all key operational variables that can be adjusted in the system, such as refrigerant circulation rate, inlet and outlet temperature settings, acid and alkali dosage ratios, heat exchanger operating time intervals, and upper and lower pH adjustment thresholds. These variables constitute a multidimensional decision space. To efficiently search for solutions across complex, multidimensional, and nonlinear objective functions and parameter spaces, a non-dominated sorting genetic algorithm (NSA) was introduced as a solver. This algorithm simulates the "selection, crossover, and mutation" mechanisms of biological evolution, selecting mutually non-dominated solutions from each generation of candidate solutions. This algorithm then iteratively optimizes the solution to obtain a Pareto optimal solution set. Each solution in this set represents a parameter combination that achieves the best compromise between energy efficiency, acid water treatment efficiency, and operating costs. The entire set describes the optimal operating boundary of the system under different priorities. Each candidate solution in the Pareto solution set is evaluated for satisfaction. Fuzzy membership function technology is employed. This technique sets an ideal value and an acceptable threshold for each objective function. Based on the deviation between the target value and the ideal value, the membership value of each solution in that objective dimension is calculated. The membership values ​​across all objective dimensions are weighted and aggregated to produce an overall satisfaction score. The control parameter combination with the highest satisfaction score among all Pareto solutions is selected as the optimal control solution for the current operating conditions, ensuring its comprehensive adaptability to system objective satisfaction. The optimal control parameter combination is obtained to enhance the system's ability to quickly respond to and accurately adjust the corrosion state, and the pH adjustment process is controlled in stages.The pH adjustment process is divided into two stages: coarse adjustment and fine adjustment. The coarse adjustment stage is used to quickly correct the pH deviation value. When the system detects that the pH value is far away from the target range, a large adjustment coefficient is activated to speed up the reaction speed and quickly return the pH value to the target critical area. When the pH value approaches the target range and enters the fine adjustment stage, the adjustment coefficient is scaled and a smaller adjustment amplitude is used to achieve fine adjustment, thereby avoiding over-adjustment or oscillation. The adjustment amplitude setting of the entire staged adjustment strategy is dynamically adjusted based on the pH adjustment coefficient obtained by the previous calculation. The coefficient is adaptively updated according to the system response sensitivity, adjustment inertia and disturbance frequency to ensure that the adjustment process has both response speed and adjustment accuracy. Based on the above-mentioned linkage control logic of coarse adjustment and fine adjustment, a set of anti-corrosion control strategies that adapt to the current corrosion risk level, energy efficiency goals and economic constraints are formed.

[0029] In a specific embodiment, the direct-spray waste heat recovery acidic water treatment method further includes the following steps: Obtain real-time status parameters including temperature, pressure, flow, and acidity from the target monitoring points of the buried pipe cold storage system, preprocess and normalize the real-time status parameters, and obtain the current operating status matrix of the system; The upper-layer agent input vector is constructed based on the system's current operating state matrix and the control parameters in the anti-corrosion control strategy. The target features are extracted through the feature selection algorithm to obtain the agent-layer decision data. The upper-level decision function is applied to the agent-level decision data to perform a global strategy analysis to obtain the system adjustment direction. The execution-level optimization target is determined based on the system adjustment direction to obtain the direct-spray waste heat recovery control parameters.

[0030] Specifically, real-time status data, including key indicators such as temperature, pressure, flow rate, and acidity, is collected from target monitoring points within the underground pipe thermal storage system. This data provides information about system operation and is fundamental for evaluating and optimizing system performance. Real-time monitoring of these status parameters allows for timely understanding of the system's operating status and identification of potential operational issues, such as excessive temperature, unstable flow rate, or excessive acidity. To ensure the accuracy and effectiveness of subsequent analysis, the real-time status parameters undergo preprocessing and normalization. This preprocessing includes steps such as outlier detection and removal, data smoothing, and missing value filling to ensure data integrity and consistency. Normalization converts each parameter to the same dimensional range. For example, scaling temperature, pressure, flow rate, and acidity values ​​to the standardized interval [0, 1] ensures that parameters of different scales can be effectively compared and processed within the same data framework. After preprocessing and normalization, the system's current operating status matrix is ​​obtained. This matrix records the status information of each monitoring point at a specific moment and reflects the system's real-time operating status. Based on the system's current operating state matrix and the control parameters of the anti-corrosion control strategy, an upper-level agent input vector is constructed. This vector includes various system characteristics extracted from the state matrix, such as temperature change rate, pressure fluctuation, flow rate variation, and dynamic changes in acidity. Furthermore, control parameters from the anti-corrosion control strategy, such as pH adjustment coefficient, flow rate adjustment range, and temperature control target, are integrated into the input vector. This input vector reflects the system's physical state and includes essential information for anti-corrosion control. To extract the features most influential on system regulation, a feature selection algorithm is used to identify the most representative and predictive features from the input features, while removing redundant or irrelevant features to reduce computational complexity and improve decision accuracy. Feature selection methods include information gain, variance analysis, and principal component analysis. These methods automatically select the most valuable features for system regulation based on data distribution and correlation. Through feature selection, decision data for the agent layer is generated. This data is then fed into the upper-level decision function for global policy analysis. The upper-level decision function comprehensively considers all system input data and control objectives, and uses mathematical models and optimization algorithms to determine the system's adjustment direction. For example, based on the changing trends of real-time state parameters, it analyzes whether the current system operation is trending towards an abnormal state or deviating from optimal operating conditions, and accordingly determines whether adjustments to temperature, flow rate, or pH value are necessary. This decision function, based on technologies such as multi-objective optimization algorithms, machine learning models, or rule engines, dynamically evaluates the current state of the system and provides reasonable adjustment suggestions. Through global strategy analysis, it is clear in which direction the system should be adjusted to optimize heat transfer, reduce corrosion risks, and ensure the long-term stable operation of the system. Based on the output of the upper-level decision function, that is, the direction of system adjustment, the optimization objectives of the execution layer are determined.The optimization objectives of the execution layer include several aspects, such as optimizing the system's energy consumption, improving heat recovery efficiency, ensuring that the pH value remains within a reasonable range, and ensuring stable flow and pressure. Based on the system's adjustment direction, each control variable is optimized, such as adjusting the flow rate based on heat exchange efficiency requirements, adjusting the pH value based on corrosion protection requirements, and adjusting the equipment load based on temperature changes. These adjustments ensure optimal system performance under different operating conditions, while also reducing corrosion risks and extending equipment life. Based on the optimal control parameter combination, efficient operation and corrosion protection control are achieved in the direct-injection waste heat recovery process. Optimizing the control parameter combination requires comprehensive consideration of various objective functions, such as energy utilization efficiency, acidic water treatment efficiency, and system operating costs, to ensure that the system's economy and efficiency are maximized while meeting corrosion protection requirements. By gradually optimizing and adjusting system parameters in real time, intelligent system management and refined regulation are achieved, improving overall performance and safety.

[0031] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0032] If the integrated unit is implemented as a software functional unit and sold or used as a standalone product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product, stored in a storage medium, includes instructions for causing a direct-spray waste heat recovery acidic water treatment device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0033] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A direct shower waste heat recovery acidic water treatment method, characterized in that: include: A pH differential sensor group and a conductivity sensor are installed at the interface between the buried pipe cold storage system and the direct-spray waste heat recovery device to monitor the entire acidic water treatment process in real time and obtain data on changes in acidic water quality. Filtering and analyzing the acidic water quality change data and system operating parameters and operating condition matching to obtain a cold storage-acidic water treatment synergistic model; The cold storage-acid water treatment synergistic model and underground cold storage state data are input into a hyperbolic mapping network for thermochemical coupling analysis to obtain corrosion risk assessment results; A dual-constraint optimization model of heat exchange efficiency and pH value adjustment is executed according to the corrosion risk assessment result, and an anti-corrosion control strategy is output.

2. The direct-spray waste heat recovery acidic water treatment method according to claim 1, characterized in that: A pH differential sensor group and a conductivity sensor are installed at the interface between the buried pipe cold storage system and the direct shower waste heat recovery device to monitor the entire acidic water treatment process in real time and obtain acidic water quality change data, including: A first pH sensor and a first conductivity sensor are installed between the water outlet of the buried pipe cold storage system and the water inlet of the direct-spray waste heat recovery device to establish an inlet parameter monitoring point and obtain acidic water inlet characteristic data; A second pH sensor and a second conductivity sensor are installed between the water outlet of the direct-spray waste heat recovery device and the inlet of the acidic water collection tank to establish an outlet water parameter monitoring point and obtain acidic water outlet characteristic data; Performing differential calculation on the acidic water inlet characteristic data and the acidic water outlet characteristic data to obtain a pH value difference ΔpH and an electrical conductivity difference ΔEC value; Correlating the pH value difference ΔpH and the conductivity difference ΔEC with the liquid level sensor and temperature sensor data in the acidic water collection tank to obtain a multidimensional water quality parameter matrix; Arranging the multidimensional water quality parameter matrix according to a time series and applying a sliding window algorithm to process it to obtain dynamic water quality change trend data; The dynamic water quality change trend data is compared and analyzed with a preset acidic water treatment standard range, and abnormal points are marked to obtain acidic water quality change data.

3. The direct-spray waste heat recovery acidic water treatment method according to claim 1, characterized in that: The filtering and operating condition matching analysis of the acidic water quality change data and system operating parameters is performed to obtain a cold storage-acidic water treatment collaborative mode, including: The pH value and conductivity in the acidic water quality change data are reorganized into time series to obtain an initial water quality feature sequence, and the temperature sensor and pressure sensor in the buried pipe cold storage system are preprocessed in real time to obtain a system operation feature sequence; Merging the initial water quality characteristic sequence with the system operation characteristic sequence to obtain an original coupling data matrix; Adopting an adaptive Kalman filter algorithm to dynamically suppress noise on the original coupled data matrix to obtain a low-noise data stream; Constructing a multivariable transfer function based on the temporal relationship between the pH value and the system temperature and pressure changes in the low-noise data stream to obtain an acid-base-temperature-pressure response model; Performing parameter identification and sensitivity analysis on the acid-base-temperature-pressure response model, extracting target influencing factors and their quantitative relationships, and obtaining acid-base balance correlation data; The refrigerant-acidic water heat exchange index and pH value adjustment coefficient are calculated based on the acid-base balance correlation data, and an operating condition matching analysis is performed to obtain a cold storage-acidic water treatment synergistic mode.

4. The direct-spray waste heat recovery acidic water treatment method according to claim 3, characterized in that: The refrigerant-acidic water heat exchange index and pH value adjustment coefficient are calculated based on the acid-base balance correlation data, and a working condition matching analysis is performed to obtain a cold storage-acidic water treatment synergistic mode, including: Extracting the temperature difference, pH value change rate, and conductivity change value from the acid-base balance correlation data, and performing heat exchange efficiency calculation based on the temperature difference, pH value change rate, and conductivity change value to obtain a refrigerant-acid water heat exchange index; performing adaptive adjustment according to the deviation between the current pH value and the target pH value in the acid-base balance associated data to obtain a pH value adjustment coefficient; The refrigerant-acid water heat exchange index, the pH value adjustment coefficient and the system operating state parameters are combined into a working condition matching matrix, and a fuzzy clustering analysis is performed on the working condition matching matrix to obtain a system operating condition classification; A state transition probability matrix is ​​constructed based on the system operating condition classification, and an optimal strategy function is calculated by applying a Markov decision process to obtain a mode switching decision table; The mode switching decision table is used to perform parameter matching calculation with the refrigerant-acid water heat exchange index and the pH value adjustment coefficient to obtain a cold storage-acid water treatment coordinated mode.

5. The direct-spray waste heat recovery acidic water treatment method according to claim 1, characterized in that: The cold storage-acid water treatment synergistic mode and underground cold storage state data are input into a hyperbolic mapping network for thermochemical coupling analysis to obtain corrosion risk assessment results, including: Extracting the operating condition type parameters, heat exchange performance indexes and pH adjustment quantitative values ​​as collaborative mode characteristic data according to the cold storage-acid water treatment collaborative mode; The soil temperature distribution and heat flux density data are collected from the temperature sensor at the buried location of the buried pipe-in-pipe heat exchanger as underground cold storage status data; Inputting the collaborative mode feature data and the underground cold storage state data into a hyperbolic embedding layer of a hyperbolic surface mapping network, and mapping the Euclidean space features to a hyperbolic space through the hyperbolic embedding layer to obtain a nonlinear dimensionality reduction representation; Inputting the nonlinear dimensionality reduction representation into a thermochemical coupling layer of a hyperbolic mapping network, performing a hyperbolic convolution operation on the thermochemical coupling layer to extract the interaction characteristics between pH value changes and temperature gradients to obtain corrosion driving factors; Inputting the corrosion driving factor into the risk quantification layer of the hyperbolic mapping network, calculating the electrochemical reaction activity of the material-environment interface through the risk quantification layer to obtain a spatially discretized corrosion rate distribution; The corrosion rate distribution is extrapolated in time series and compared with the material corrosion resistance threshold. The risk level is calculated using a hyperbolic distance metric function to obtain a corrosion risk assessment result.

6. The direct-spray waste heat recovery acidic water treatment method according to claim 1, characterized in that: The dual-constraint optimization model for heat exchange efficiency and pH value adjustment is executed according to the corrosion risk assessment result to output an anti-corrosion control strategy, including: Converting the risk level in the corrosion risk assessment result into a corrosion protection constraint condition; Based on the corrosion protection constraints, an optimization target model is constructed, which includes an energy utilization efficiency objective function, an acidic water treatment efficiency objective function, and a system operation cost objective function; Defining a control parameter space according to the optimization target model, and performing a non-dominated sorting genetic solution based on the control parameter space and the optimization target model to obtain a Pareto optimal solution set; Performing a satisfaction evaluation on each solution in the Pareto optimal solution set using a fuzzy membership function to obtain a satisfaction score, and selecting an optimal control parameter combination according to the satisfaction score; Based on the optimal control parameter combination, the pH value adjustment process is divided into two stages: coarse adjustment and fine adjustment. Different adjustment ranges are designed in combination with the pH adjustment coefficient, and an anti-corrosion control strategy is generated.

7. The direct shower type waste heat recovery acidic water treatment method according to claim 1, characterized in that: The direct-spray waste heat recovery acidic water treatment method further comprises: Acquire real-time status parameters including temperature, pressure, flow, and acidity from target monitoring points of the buried pipe cold storage system, and preprocess and normalize the real-time status parameters to obtain the current operating status matrix of the system; Constructing an upper-layer agent input vector based on the current operating state matrix of the system and the control parameters in the anti-corrosion control strategy, and extracting target features through a feature selection algorithm to obtain agent-layer decision data; Applying the upper-layer decision function to the agent-layer decision data to perform global strategy analysis, obtain the system adjustment direction, and determine the execution-layer optimization target according to the system adjustment direction to obtain the direct-spray waste heat recovery control parameters.

Images