A comprehensive seawater resource utilization system and method based on multi-energy complementarity and intelligent regulation

Abstract

This invention relates to the field of comprehensive utilization and intelligent control of seawater resources, specifically to a comprehensive utilization system and method for seawater resources based on multi-energy complementarity and intelligent regulation. The method collects operational status data and external environmental data from energy supply, seawater desalination, and chemical resource extraction processes, constructs a global state vector and a dynamic mathematical model, and uses a model predictive control algorithm to continuously solve and optimize the control instruction set. Based on this, it coordinates and controls the load allocation of the seawater desalination process and the process parameters of the chemical resource extraction process. When product purity is abnormal, it reverse-tracks and corrects the desalination operating parameters, and performs online identification and updates when the deviation of the comprehensive benefit index continuously exceeds a threshold. This invention can achieve synergistic optimization of multi-energy supply and multiple processes, improving the comprehensive utilization efficiency, operational stability, and overall benefits of water and chemical resources in seawater.

Description

Technical Field

[0001] This invention relates to the field of comprehensive utilization and intelligent control of seawater resources, specifically to a comprehensive utilization system and method for seawater resources based on multi-energy complementarity and intelligent regulation. Background Technology

[0002] Comprehensive utilization of seawater resources is an important technological direction for alleviating freshwater shortages and improving the efficiency of marine resource development. Existing seawater utilization systems typically focus not only on freshwater production but also on the extraction and utilization of chemical resources such as bromine, magnesium, potassium, and lithium from concentrated brine to improve overall resource utilization and economic benefits. Meanwhile, with the development of solar energy, waste heat utilization, energy storage devices, and grid-coordinated energy supply technologies, the energy supply methods involved in seawater desalination and chemical resource extraction are becoming increasingly diverse, and system operation exhibits significant characteristics of multi-energy coupling, multi-process series connection, and multi-objective optimization.

[0003] However, in existing technologies, the seawater desalination and chemical resource extraction processes often employ relatively independent control methods. The lack of unified coordination between the energy supply side and the process operation side makes it difficult to coordinate and adjust low-temperature multi-effect distillation, reverse osmosis, and subsequent resource extraction processes in response to external energy fluctuations and changes in operating conditions. Especially under conditions of varying solar irradiance, fluctuating time-of-use electricity prices, unstable waste heat input, and fluctuating brine concentrations, existing control methods typically only allow for localized adjustments to a single unit or process, making it difficult to simultaneously consider water production demand, energy costs, resource extraction revenue, and product quality stability. Furthermore, when abnormal product purity occurs during the chemical resource extraction process, existing systems typically lack a closed-loop control mechanism that traces back from the extraction end to the desalination end and promptly corrects operating parameters, and also lack the ability to update model parameters online based on actual operational deviations.

[0004] Therefore, existing technologies still suffer from problems such as insufficient coordination between multiple energy supply and multiple process flows, weak coupling control between desalination and extraction, lack of reverse correction under abnormal operating conditions, and poor model adaptability, making it difficult to achieve the coordinated and efficient utilization of water and chemical resources in seawater. To address this, it is necessary to provide a comprehensive seawater resource utilization method that improves the overall system efficiency, operational stability, and resource utilization efficiency by unifying modeling, continuously optimizing, and coordinating the energy supply, seawater desalination, and chemical resource extraction processes. Summary of the Invention

[0005] The technical problem this invention aims to solve is to address the shortcomings of existing technologies by providing a comprehensive seawater resource utilization system and method based on multi-energy complementarity and intelligent regulation. This method collects operational status data and external environmental data from energy supply, seawater desalination, and chemical resource extraction processes, constructs a global state vector and a dynamic mathematical model, and employs a model predictive control algorithm to continuously optimize the control instruction set. Based on this, it coordinates and controls the load allocation of the seawater desalination process and the process parameters of the chemical resource extraction process. When product purity is abnormal, it reverse-tracks and corrects the desalination operating parameters, and performs online identification and updates when the deviation of the comprehensive benefit index continuously exceeds a threshold. This invention enables the synergistic optimization of multi-energy supply and multiple process flows, improving the comprehensive utilization efficiency, operational stability, and overall benefits of water and chemical resources in seawater.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A comprehensive utilization method for seawater resources based on multi-energy complementarity and intelligent regulation, the method comprising:

[0008] The system collects operational status data and external environmental data of energy supply, seawater desalination and chemical resource extraction processes, and constructs a global state vector based on these data. The external environmental data includes solar irradiance prediction sequences and time-of-use electricity price sequences.

[0009] Based on the global state vector, a dynamic mathematical model of the energy supply, seawater desalination and chemical resource extraction process is constructed, and the model predictive control algorithm is used to solve the optimization control instruction set in a rolling manner.

[0010] According to the optimized control instruction set, the load allocation of the seawater desalination process and the process parameters of the chemical resource extraction process are coordinated and controlled.

[0011] When any step in the chemical resource extraction process results in abnormal product purity, the process traces back to the source and corrects the operating parameters of the seawater desalination process, forming a corrective closed loop.

[0012] When the deviation between the comprehensive benefit index obtained from actual operation and the comprehensive benefit index predicted by the dynamic mathematical model continues to exceed a set threshold, the dynamic mathematical model is identified and updated online based on the actual operation data to achieve the synergistic and efficient utilization of water resources and chemical resources in seawater.

[0013] Furthermore, the construction of the dynamic mathematical model and the rolling solution optimization of the control instruction set includes:

[0014] Based on the global state vector, an energy supply sub-model, a seawater desalination sub-model, and a chemical resource extraction sub-model are constructed respectively to form the dynamic mathematical model.

[0015] Under the premise of satisfying the process constraints and equipment operation constraints of each process, the comprehensive objective function consisting of energy consumption cost and comprehensive resource extraction benefits is used as the optimization objective, and the optimization control instruction set is solved in a rolling manner. The optimization control instruction set includes the target load of low temperature multi-effect distillation process, the target load of reverse osmosis process, the target power of microwave heat source, the charging and discharging power of energy storage device, and the amount of reagents added in each process of chemical resource extraction.

[0016] Furthermore, the energy supply sub-model, seawater desalination sub-model, and chemical resource extraction sub-model respectively include:

[0017] The energy supply sub-model uses the energy supply parameters of solar thermal collectors, grid power supply, microwave heat source, energy storage device charging and discharging parameters, and power plant waste heat as inputs to establish dynamic energy balance equations for the energy load of each process flow.

[0018] The seawater desalination sub-model uses the heat source parameters of the low-temperature multi-effect distillation process and the operating pressure of the reverse osmosis process as inputs to calculate the flow rate and concentration of the concentrated brine output from the low-temperature multi-effect distillation process and the reverse osmosis process, respectively, and calculates the concentration of the concentrated brine after mixing accordingly.

[0019] The chemical resource extraction sub-model takes the feed ion concentration and reagent dosage of the evaporation crystallization, bromine extraction, magnesium / potassium extraction and lithium extraction processes as inputs, calculates the product yield of each process, and sequentially transfers the ion concentration in the output mother liquor of each process to the next process as the feed ion concentration.

[0020] Furthermore, the rolling solution optimization control instruction set includes:

[0021] The process constraints and equipment operation constraints of each process flow are used as the solution conditions. The process constraints include the water production rate not being lower than the preset demand, the upper limit of the temperature of the low temperature multi-effect distillation process, the operating pressure range of the reverse osmosis process, and the target range of feed ion concentration for each process. The equipment operation constraints include the upper and lower limits of the load of each process flow, the state of charge range of the energy storage device, and the rate of change of action of each actuator.

[0022] Within each control cycle, using the current global state vector as the initial condition, the future operating state of each process is predicted based on the dynamic mathematical model, and the optimized control instruction set is solved.

[0023] The control instructions corresponding to the current control cycle are executed from the optimized control instruction set, and the solution is recalculated in the next control cycle with the updated global state vector as the initial condition to achieve rolling optimization.

[0024] Furthermore, the coordination control includes:

[0025] Based on the solar irradiance prediction sequence and the time-of-use electricity price sequence, the current operating scenario is determined, and the system switches between high heat energy supply scenario, low heat energy off-peak electricity price scenario and low heat energy peak electricity price scenario to dynamically adjust the load distribution of the low temperature multi-effect distillation process and the reverse osmosis process.

[0026] The change in brine concentration in the next control cycle is predicted based on the dynamic mathematical model. The change in brine concentration and the predicted concentration of each ion are then fed forward to the control unit of each chemical resource extraction process to control the dynamic adjustment of reagent dosage according to the feed ion concentration, thus forming a cascade parameter linkage control.

[0027] Furthermore, the dynamic adjustment of the load distribution between the low-temperature multi-effect distillation process and the reverse osmosis process includes:

[0028] In high heat energy supply scenarios, when the solar thermal power is not lower than the rated heat load of the low temperature multi-effect distillation process, the operating load of the low temperature multi-effect distillation process is increased to prioritize the consumption of solar thermal power, the operating load of the reverse osmosis process is reduced accordingly, and the excess heat energy is stored in the thermal storage device.

[0029] In scenarios with low heat energy and low off-peak electricity prices, when the solar thermal power is lower than the heat demand corresponding to the current target load of the low-temperature multi-effect distillation process, and the time-of-use electricity price is not higher than the preset peak electricity price threshold, the reverse osmosis process load is increased to use low-priced electricity to produce water, and at the same time, the microwave heat source is activated to supplement the heat required by the low-temperature multi-effect distillation process to maintain its minimum operating load.

[0030] In scenarios with low heat energy and peak electricity price, when the solar thermal power is lower than the heat demand corresponding to the current target load of the low temperature multi-effect distillation process, and the time-of-use electricity price is higher than the preset peak electricity price threshold, the low temperature multi-effect distillation process and the reverse osmosis process are simultaneously reduced to the minimum operating load, and the energy storage device is called to release electrical energy to supplement the electricity required for basic water production.

[0031] When the waste heat supply from the power plant reaches the preset waste heat access threshold, the waste heat from the power plant is identified as the priority heat source for the low-temperature multi-effect distillation process. The output of the solar thermal collector is reduced and the output of the microwave heat source is lowered. Based on the waste heat from the power plant, the target load of the low-temperature multi-effect distillation process is recalculated, and the target load of the reverse osmosis process is reduced accordingly.

[0032] Furthermore, the cascaded parameter linkage control includes:

[0033] Based on the change in the concentration of the concentrated brine, the feed rate and evaporation load of the evaporation and crystallization process are adjusted, and the predicted concentrations of each ion in the output mother liquor are sequentially fed forward to the bromine extraction, magnesium / potassium extraction and lithium extraction processes.

[0034] The bromine extraction process dynamically adjusts the oxidant dosage based on the bromide ion concentration in the feed, and transfers the ion concentration in the mother liquor after bromine extraction to the magnesium / potassium extraction process.

[0035] The magnesium / potassium extraction process dynamically adjusts the amount of precipitant added based on the magnesium and potassium ion concentrations in the feed, and potassium extraction is carried out after the magnesium removal meets the standard. The ion concentration in the mother liquor after magnesium / potassium extraction is then transferred to the lithium extraction process.

[0036] The lithium extraction process dynamically adjusts the amount of adsorbent and the elution frequency based on the lithium ion concentration and magnesium-lithium ratio in the feed.

[0037] Furthermore, the process of reverse tracing and correcting the operating parameters of the seawater desalination process includes:

[0038] When the purity of the product in any process is lower than the set purity threshold, the feed ion concentration of that process is checked to see if it deviates from the preset normal concentration range.

[0039] If the feed ion concentration of this process does not deviate from the preset normal concentration range, it is determined that the process is abnormal and the dosage of the reagent in this process is adjusted.

[0040] If the feed ion concentration of this process deviates from the preset normal concentration range, it is determined that the raw material quality is abnormal, and an abnormal signal is fed back to the model prediction control algorithm.

[0041] The model predictive control algorithm increases the concentration of the brine output from the seawater desalination process based on the abnormal signal, thereby increasing the concentration of the target ions to be extracted.

[0042] The corrected brine concentration prediction is fed forward to the control units of each chemical resource extraction process to trigger a readjustment of the reagent dosage in each process, thus forming a corrective closed loop.

[0043] Furthermore, the online identification and updating of the dynamic mathematical model based on actual operational data includes:

[0044] The number of control cycles in which the deviation exceeds a set threshold is counted. When the number of control cycles reaches a preset threshold for consecutive cycles, an online identification update is triggered.

[0045] Collect actual operating data within a preset quantity control period before the triggering time, and identify and update the key parameters of the energy supply sub-model, the seawater desalination sub-model, and the chemical resource extraction sub-model respectively;

[0046] The updated model parameters are written into the model predictive control algorithm and take effect from the next control cycle.

[0047] A comprehensive seawater resource utilization system based on multi-energy complementarity and intelligent regulation, the system comprising:

[0048] The data acquisition unit is used to collect operational status data and external environmental data of energy supply, seawater desalination and chemical resource extraction processes, wherein the external environmental data includes solar irradiance prediction sequences and time-of-use electricity price sequences.

[0049] A state construction unit is used to construct a global state vector based on the running state data and external environment data;

[0050] The dynamic modeling and optimization unit is used to construct dynamic mathematical models of energy supply, seawater desalination and chemical resource extraction processes based on the global state vector, and to solve the optimized control instruction set in a rolling manner.

[0051] The coordination and control unit is used to coordinate and control the load distribution of the seawater desalination process and the process parameters of the chemical resource extraction process according to the optimized control instruction set.

[0052] The anomaly correction unit is used to reverse the source and correct the operating parameters of the seawater desalination process when an abnormality in product purity occurs at any step in the chemical resource extraction process, forming a closed loop of correction.

[0053] The online update unit is used to identify and update the dynamic mathematical model online based on actual operating data when the deviation between the comprehensive benefit index obtained from actual operation and the comprehensive benefit index predicted by the dynamic mathematical model continues to exceed a set threshold.

[0054] Compared with the prior art, the beneficial effects of the present invention are:

[0055] This invention collects operational status data and external environmental data from energy supply, seawater desalination, and chemical resource extraction processes, constructs a global state vector and a dynamic mathematical model, and employs a model predictive control algorithm to continuously optimize the control instruction set. This achieves coordinated optimization control between the multi-energy supply side and multiple processes such as seawater desalination and chemical resource extraction. As a result, it can dynamically adjust the load distribution of the low-temperature multi-effect distillation process and the reverse osmosis process according to changes in solar irradiance, time-of-use electricity price fluctuations, and waste heat access, while coordinating the process parameters of each extraction step. This improves the comprehensive utilization efficiency of water and chemical resources in seawater and reduces operating energy consumption and overall costs. Attached Figure Description

[0056] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0057] Figure 1A schematic diagram of the structure and workflow of a comprehensive seawater resource utilization system provided in an embodiment of the present invention;

[0058] Figure 2 A flowchart illustrating the comprehensive utilization method of seawater resources provided in this embodiment of the invention.

[0059] Figure 3 This is a schematic diagram of the structure of a comprehensive seawater resource utilization system provided in an embodiment of the present invention. Detailed Implementation

[0060] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0061] Example 1

[0062] In coastal industrial parks, island water supply stations, and port-adjacent resource utilization facilities, seawater is often no longer used solely as a desalination feedstock, but rather serves the dual purpose of freshwater supply and concentrated brine extraction. Especially in scenarios involving the integrated utilization of multiple energy sources such as solar thermal collectors, grid power, energy storage devices, industrial waste heat, or microwave heat sources, the seawater desalination section and the subsequent chemical resource extraction section exhibit a clear coupling relationship in terms of material flow, energy flow, and profit objectives: changes in the front-end desalination load directly affect the concentration and ionic composition of the downstream concentrated brine, while changes in the product purity and yield of the downstream extraction process reflect the stability of the front-end supply. Therefore, the solution described in this application is applicable to integrated seawater resource utilization scenarios with fluctuating multi-energy supply, tandem operation of desalination and extraction, and the need for synergistic optimization of comprehensive benefits, rather than based on a fixed process layout, a single energy type, or a single extraction target.

[0063] In long-term engineering practice in this field, existing systems are typically controlled separately for three stages: energy supply, seawater desalination, and chemical extraction. The energy side focuses on independent scheduling, the desalination side operates based on predetermined loads or empirical parameters, and the extraction side mainly makes local adjustments based on current feed conditions. This type of control can maintain basic operation when conditions are stable, but when there are fluctuations in solar irradiance, time-of-use electricity price changes, changes in waste heat input, or deviations in the composition of concentrated brine, information transmission and adjustment responses between upstream and downstream processes often lag. This can easily lead to unreasonable desalination load allocation, deviations in reagent dosing in subsequent processes, increased fluctuations in product purity, and an imbalance between overall energy consumption and revenue. In particular, when purity anomalies occur at the extraction end, existing technologies can usually only perform local compensation within the abnormal process, making it difficult to trace back to the upstream desalination operation status and implement effective corrections. This results in the passive amplification of anomalies and insufficient system adaptability.

[0064] Please see Figure 1Based on the above understanding, this application does not treat seawater desalination and chemical resource extraction as independent, sequential units. Instead, it constructs a dynamic mathematical model for the entire process, starting from the overall relationship that "energy supply determines desalination load, desalination conditions determine brine quality, brine quality determines extraction effect, and extraction anomalies can inversely characterize desalination deviations." A rolling optimization and feedback correction mechanism is introduced. On this basis, on the one hand, the low-temperature multi-effect distillation, reverse osmosis, and subsequent extraction processes are coordinated and controlled according to changes in the external environment and system operating status. On the other hand, when product purity is abnormal or the overall benefit deviation continues to expand, the adaptability of the model and control strategy to actual operating conditions is improved through reverse tracing and online identification and updating. Therefore, this application aims to solve not the control problem of a single device, but the problem of coordinated control across processes, energy sources, and objectives in a comprehensive seawater resource utilization system under multi-energy complementarity conditions.

[0065] Please see Figure 2 The present invention provides an embodiment of a comprehensive utilization method for seawater resources based on multi-energy complementarity and intelligent regulation. The specific steps of the method are as follows:

[0066] S1: Collect operational status data and external environmental data of energy supply, seawater desalination and chemical resource extraction processes, and construct a global state vector accordingly. The external environmental data includes solar irradiance prediction sequences and time-of-use electricity price sequences.

[0067] In this embodiment, this step is preferably applied to integrated utilization scenarios that simultaneously configure solar thermal collection units, grid power supply units, energy storage units, and seawater desalination and brine extraction devices, such as island desalination stations, port-adjacent resource utilization devices, or coastal parks with industrial waste heat coupling conditions. The operational status data is not limited to a fixed format. Those skilled in the art will understand that it is sufficient as long as it can at least reflect the current operational status of each process segment and support subsequent modeling and control. Specifically, it may include heat supply, power supply, and energy storage status on the energy side; distillation temperature, operating pressure, water production, brine flow rate and concentration on the desalination side; and feed ion concentration, reagent dosage, process dwell time, and product purity on the extraction side. External environmental data should at least include solar irradiance prediction sequences and time-of-use electricity price sequences. When waste heat access conditions exist, it may also include information on available waste heat. By unifying the aforementioned multi-source heterogeneous data into a global state vector, information originally scattered across energy supply, seawater desalination, and chemical extraction can be mapped onto the same decision-making basis. This avoids the information fragmentation problem caused by each process segment collecting and judging its own data in existing technologies, and provides a unified state entry point for subsequent cross-process coupling modeling and rolling optimization. Its direct effect is that it can simultaneously characterize energy supply conditions, desalination side operating conditions, and extraction side feed quality changes.

[0068] S2: Based on the global state vector, construct a dynamic mathematical model of the energy supply, seawater desalination and chemical resource extraction process, and use a model predictive control algorithm to solve and optimize the control instruction set.

[0069] In this embodiment, the dynamic mathematical model does not model a single device in isolation, but rather establishes separate sub-models for energy supply, seawater desalination, and chemical resource extraction, forming a unified process correlation through the relationship between energy flow and material flow. Specifically, the energy supply sub-model describes the energy balance between solar energy, the power grid, energy storage, microwave heat sources, and waste heat; the seawater desalination sub-model characterizes the water production and brine output characteristics of low-temperature multi-effect distillation and reverse osmosis under different heat source and pressure conditions; and the chemical resource extraction sub-model characterizes the yield and purity variations of processes such as evaporation crystallization, bromine extraction, magnesium / potassium extraction, and lithium extraction as the feed composition changes. The model predictive control algorithm preferably executes in a rolling manner with a set control cycle. Within each cycle, using the aforementioned global state vector as initial conditions, it predicts the operating state over several future time domains. Under constraints such as water production demand, temperature and pressure boundaries, ion concentration range, energy storage state of charge, and actuator action rate, it solves for an optimized control instruction set including distillation load, reverse osmosis load, microwave heat source power, energy storage charging and discharging power, and reagent dosage. The purpose of this approach is to transform the static scheduling method that originally relied on experience into predictive decision-making oriented towards future states, so that energy utilization, desalination operation and revenue extraction can be considered in a unified optimization framework. Its advantage is that it can take into account water production, cost, revenue and stability in scenarios with large fluctuations in external energy and changes in operating conditions, and reduce the chain deviations caused by local optimal control.

[0070] S3: Based on the optimized control instruction set, coordinate and control the load allocation of the seawater desalination process and the process parameters of the chemical resource extraction process.

[0071] In this embodiment, the objects of coordinated control include at least the low-temperature multi-effect distillation process and reverse osmosis process on the desalination side, and the evaporation crystallization, bromine extraction, magnesium / potassium extraction, and lithium extraction processes on the extraction side. Specifically, when solar energy supply is strong, the low-temperature multi-effect distillation load is prioritized to fully utilize thermal energy; when solar energy is insufficient and electricity prices are low, the reverse osmosis load is appropriately increased and supplemental heating is used to maintain the minimum operating requirements of the distillation section; when solar energy is insufficient and electricity prices are high, the loads of the two desalination processes are reduced and energy storage is used to release electrical energy to maintain basic water production capacity. Simultaneously with changes in the load distribution on the desalination side, based on the predicted changes in concentrated brine concentration and the predicted concentrations of each ion obtained from the dynamic mathematical model, feedforward adjustments are made to the reagent dosage, evaporation load, feed rate, and subsequent separation parameters of each process on the extraction side. This ensures that the downstream processes no longer passively wait for anomalies to occur before making corrections, but instead adjust operating parameters in advance based on changes in the composition of the upstream materials. Those skilled in the art will understand that specific control parameters can be selected according to different process layouts and target products, as long as the transmission of material state changes on the desalination side to the adjustment action on the extraction side can be achieved to a minimum. This application does not impose stricter limitations in this regard. The purpose of this step is to solve the problem of segmented control and disconnect between desalination and extraction in the prior art. Its advantage is that it can reduce the impact of concentrated brine fluctuations on the stability of subsequent extraction, and improve the continuity of the resource extraction process and the controllability of product quality.

[0072] S4: When any step in the chemical resource extraction process results in abnormal product purity, the source is traced back and the operating parameters of the seawater desalination process are corrected to form a corrective closed loop.

[0073] In this embodiment, abnormal product purity is not automatically attributed to the abnormal process itself. Instead, it is first determined whether the feed ion concentration of that process deviates from the preset normal concentration range. If it does not deviate, it is primarily identified as a process deviation in that process, such as abnormal reagent addition, residence time, or local reaction conditions. If it has deviated, the source of the abnormality is further identified as being related to the upstream feed status, and an abnormal signal is fed back to the model predictive control algorithm to trigger a reverse correction of the operating parameters on the seawater desalination side. The reverse correction is not limited to fine-tuning a single parameter; preferably, it can be reflected in increasing the concentration of the brine output from the seawater desalination process, or adjusting the load ratio of low-temperature multi-effect distillation and reverse osmosis to change the feed composition of the subsequent extraction stage. After the correction is completed, the updated predicted brine concentration value is fed forward again to each extraction process to trigger a further adjustment of process parameters such as reagent addition. The core of this step is not simply fault alarms, but establishing a causal backtracking path from abnormal behavior at the extraction end to the material supply status at the desalination end. This avoids the problem in existing technologies where compensation is only repeated within the abnormal process, but the root cause of upstream deviations cannot be eliminated. Its direct effect is to improve the speed of product purity recovery and reduce the spread of abnormalities and the accumulation of process fluctuations.

[0074] S5: When the deviation between the comprehensive benefit index obtained from actual operation and the comprehensive benefit index predicted by the dynamic mathematical model continues to exceed the set threshold, the dynamic mathematical model is identified and updated online based on the actual operation data to achieve the coordinated and efficient utilization of water resources and chemical resources in seawater.

[0075] In this embodiment, the comprehensive benefit index is preferably used to comprehensively reflect factors such as water production capacity, energy consumption cost, resource extraction revenue, and product quality stability. Its specific expression can be set according to the project's focus, for example, using a comprehensive evaluation value of revenue minus cost, or a weighted multi-indicator fusion value. Those skilled in the art can determine this according to actual needs. When the deviation between the actual operating results and the model prediction results continuously exceeds a set threshold within multiple consecutive control cycles, it can be considered that the original model parameters can no longer fully represent the current operating condition. At this time, actual operating data within a preset number of control cycles before the trigger time is collected to identify and correct energy efficiency parameters in the energy supply sub-model, water production and concentration characteristic parameters in the seawater desalination sub-model, and yield or purity response parameters in the chemical resource extraction sub-model, and the correction results are written into the subsequent rolling optimization process. The purpose of this step is to solve the problem of model mismatch caused by changes in the external environment, equipment aging, operating condition migration, or raw material fluctuations during long-term operation, so that the control strategy no longer depends on the static parameters in the initial modeling stage, but can be continuously updated with the actual operating state. Its advantages lie in maintaining prediction accuracy and control effectiveness, avoiding the gradual disconnect between the model and actual operation, thereby maintaining the long-term stability and comprehensive benefits of the integrated utilization process of water and chemical resources in seawater.

[0076] In the comprehensive utilization of seawater resources, although the energy supply unit, seawater desalination unit, and chemical resource extraction unit can be arranged sequentially in the process flow, they are not independent in their actual operating mechanisms. Changes in the upstream energy supply conditions not only affect the heat input, power consumption, and operating load of the desalination section, but also further affect the output flow rate, concentration degree, and ionic composition of the concentrated brine. Changes in the composition of the concentrated brine directly affect the feed conditions of subsequent processes such as evaporation and crystallization, bromine extraction, magnesium / potassium extraction, and lithium extraction, causing changes in reagent dosage, reaction intensity, and separation rhythm. In other words, in this type of process system, the operating status of any link is not only reflected in the parameter changes within that link, but is continuously transmitted along the material and energy chains, and is reflected in subsequent processes in the form of product purity, resource yield, or overall cost. Because of this continuous transmission relationship across process segments, segmented, static, or single-objective adjustments to related processes are usually insufficient to accurately reflect the overall operating status.

[0077] Furthermore, from a process control perspective, the coupling between seawater desalination and resource extraction is not merely a simple upstream-downstream relationship where the former provides raw materials and the latter completes extraction. Rather, it manifests in the direct shaping effect of the desalination stage's output physical properties on the extraction stage's operating window. For example, when the load ratio of low-temperature multi-effect distillation and reverse osmosis is adjusted, the total amount, concentration, and ion ratio of the concentrated brine entering the extraction stage often change accordingly, leading to a shift in the effective range of operating parameters for oxidants, precipitants, and adsorbents in subsequent processes. Correspondingly, if the purity of the product deviates in subsequent processes, it does not necessarily mean the anomaly occurred in that process itself; it may also stem from a gradual shift in the feed state of the upstream desalination stage. Therefore, in this type of process, product quality is not only the final output but also an indirect representation of the upstream operating status. Based on this understanding, using downstream quality information as a basis for correcting upstream operations and using changes in upstream materials as a basis for adjusting downstream parameters essentially utilizes the inherent interrelationships within the entire process chain, rather than simply controlling multiple devices in parallel.

[0078] Based on the aforementioned process characteristics, the approach adopted in this embodiment is not to model energy dispatch, desalination operation, and resource extraction separately and then loosely piece them together. Instead, it integrates changes in the external environment, energy supply status, desalination output status, and extraction response results into a unified state expression and prediction framework. This allows the control process to revolve around the "overall operating trend over a future period" rather than the "local deviation at the current moment." Therefore, during operation, on the one hand, the desalination load can be proactively allocated based on solar irradiance, time-of-use electricity prices, energy storage status, and waste heat access. On the other hand, the extraction process parameters can be feedforward adjusted based on predicted changes in brine concentration and ion composition. When the final quality performance continuously deviates from the predicted results, the model can be further corrected through online identification, ensuring that the control basis always remains consistent with the actual operating state. With this approach, the entire process chain is no longer understood as a "sequential execution of units," but rather as a dynamic process that evolves with energy conditions, material status, and quality feedback. This is more conducive to maintaining a balance between water production capacity, resource extraction efficiency, and operational economy.

[0079] The specific steps of S2 are as follows:

[0080] S2.1: Based on the global state vector, construct the energy supply sub-model, the seawater desalination sub-model, and the chemical resource extraction sub-model respectively to form the dynamic mathematical model.

[0081] Specifically, the energy supply sub-model, seawater desalination sub-model, and chemical resource extraction sub-model respectively include:

[0082] The energy supply sub-model uses the energy supply parameters of solar thermal collectors, grid power supply, microwave heat source, energy storage device charging and discharging parameters, and power plant waste heat as inputs to establish dynamic energy balance equations for the energy load of each process flow.

[0083] The seawater desalination sub-model uses the heat source parameters of the low-temperature multi-effect distillation process and the operating pressure of the reverse osmosis process as inputs to calculate the flow rate and concentration of the concentrated brine output from the low-temperature multi-effect distillation process and the reverse osmosis process, respectively, and calculates the concentration of the concentrated brine after mixing accordingly.

[0084] The chemical resource extraction sub-model takes the feed ion concentration and reagent dosage of the evaporation crystallization, bromine extraction, magnesium / potassium extraction and lithium extraction processes as inputs, calculates the product yield of each process, and sequentially transfers the ion concentration in the output mother liquor of each process to the next process as the feed ion concentration.

[0085] Specifically, the dynamic mathematical model is broken down into three sub-models: energy supply, seawater desalination, and chemical resource extraction. This is because the energy flow, material flow, and controlled objects differ among the three. Directly using a single, holistic model would easily lead to unclear variable coupling relationships, difficulties in parameter calibration, and challenges in tracing the source of anomalies. Through hierarchical modeling, the inputs, outputs, and interrelationships of each stage can be clearly defined first, and then a continuously updatable holistic model can be formed based on a unified state vector.

[0086] In this embodiment, the energy supply sub-model preferably collects parameters such as solar thermal power, grid power supply, microwave heat source adjustable power, energy storage device state of charge and charge / discharge boundary, and power plant waste heat accessibility to describe the available range of thermal and electrical energy within the current cycle. The seawater desalination sub-model preferably describes two desalination paths: low-temperature multi-effect distillation and reverse osmosis. The low-temperature multi-effect distillation part reflects the relationship between heat source parameters and water production and brine concentration, while the reverse osmosis part reflects the relationship between operating pressure and concentrated brine flow rate and concentration. The two concentrated brine streams are mixed according to their flow rate ratio to obtain a unified feed state for the subsequent extraction stage. The chemical resource extraction sub-model is established in the order of evaporation and crystallization, bromine extraction, magnesium / potassium extraction, and lithium extraction. It correlates the feed ion concentration, reagent dosage, product yield, and mother liquor composition changes of each process and transfers the mother liquor ion concentration from the previous process to the next process.

[0087] Furthermore, the model parameters can be calibrated based on equipment design parameters, historical operating records, debugging data, or pilot-scale data. For example, the concentration characteristics of low-temperature multi-effect distillation can be calibrated according to different heat load ranges, the recovery rate and concentrate discharge characteristics of reverse osmosis can be calibrated according to different pressure ranges, and the response relationships of bromine extraction, magnesium / potassium extraction, and lithium extraction processes to reagent dosage can be calibrated according to different ion concentration ranges. Through the above processing, changes in front-end energy supply can be mapped to changes in desalination section output, and changes in desalination section output can be further mapped to changes in extraction section feed, thus providing a continuous and interpretable predictive basis for subsequent rolling optimization.

[0088] S2.2: Under the premise of satisfying the process constraints and equipment operation constraints of each process flow, the comprehensive objective function consisting of energy consumption cost and comprehensive resource extraction benefits is used as the optimization objective, and the optimization control instruction set is solved in a rolling manner. The optimization control instruction set includes the target load of low temperature multi-effect distillation process, the target load of reverse osmosis process, the target power of microwave heat source, the charging and discharging power of energy storage device, and the amount of reagents added in each process of chemical resource extraction.

[0089] Specifically, a rolling solution method is adopted because external energy conditions, desalination conditions, and feed status in the extraction section all change over time. If fixed control values ​​are used for a long period, load distribution lag, reagent dosing deviations, and a decrease in overall profitability are likely to occur. By re-predicting the operating status for several future cycles within each control cycle and re-selecting the control variables for the current cycle accordingly, the control results can continuously follow changes in operating conditions.

[0090] In this embodiment, the process constraints are preferably set as follows: the total water production is not less than the preset demand, the temperature of the low-temperature multi-effect distillation process does not exceed the upper limit of the allowable range, the operating pressure of the reverse osmosis process is within the permissible range, and the feed ion concentration of each extraction process is within the corresponding target range. The equipment operation constraints preferably include the upper and lower limits of each process load, the upper and lower limits of the state of charge of the energy storage device, and the rate of change of the actuator's single-cycle action. Those skilled in the art will understand that the above thresholds can be determined based on the equipment specifications and operating records. For example, the state of charge of the energy storage device can be set to 20% to 90%, and the single-cycle adjustment range of the actuator can be set to 5% to 15% of the rated value.

[0091] Furthermore, the comprehensive target optimization is composed of a weighted average of energy consumption cost, resource extraction revenue, and operational stability correction. The energy consumption cost may include electricity purchase cost, microwave heat source energy consumption, and energy storage depreciation cost. The resource extraction revenue may include freshwater revenue and revenue from resources such as bromine, magnesium, potassium, and lithium. The operational stability correction is used to limit abrupt changes in control quantities. For example, in a water supply priority scenario, the weight of energy consumption cost can be set to 0.45, the weight of resource extraction revenue to 0.35, and the weight of operational stability correction to 0.20; in a resource recovery priority scenario, the corresponding weights can be set to 0.35, 0.45, and 0.20. Each control cycle executes only the optimal control quantity for the current cycle, and the remaining prediction results serve as a reference for resolving the problem in the next cycle. This yields the target load for low-temperature multi-effect distillation, the target load for reverse osmosis, the target power for the microwave heat source, the charging and discharging power of energy storage, and the dosage of reagents for each process, thus ensuring consistency between energy allocation, desalination operation, and extraction regulation.

[0092] The specific steps of S2.2 are as follows:

[0093] S2.2.1: The process constraints and equipment operation constraints of each process flow are used as the solution conditions. The process constraints include the water production rate not being lower than the preset demand, the upper limit of the temperature of the low temperature multi-effect distillation process, the operating pressure range of the reverse osmosis process, and the target range of feed ion concentration for each process. The equipment operation constraints include the upper and lower limits of the load of each process flow, the state of charge range of the energy storage device, and the rate of change of action of each actuator.

[0094] Specifically, in the comprehensive utilization of seawater resources, optimizing solely based on cost or benefit can easily lead to short-term optima for a particular process segment but overall operational instability. For example, excessively high heat loads in low-temperature multi-effect distillation can increase the risk of scaling; rapid increases in reverse osmosis pressure can cause abnormal membrane pressure; or continued addition of reagents after the feed ion composition in the extraction segment deviates from the applicable window can result in fluctuations in product purity. Therefore, before solving for control commands, the process and equipment boundaries should be transformed into unified constraints to limit subsequent optimization processes to a feasible, sustainable range that meets process requirements.

[0095] In this embodiment, the constraint that the water production volume is not lower than the preset demand can be determined according to the water load plan. Preferably, the larger value between the average water consumption of the same period of the previous day and the scheduling instruction of the day is taken as the benchmark. For example, if the planned water supply for a certain period is 100 cubic meters per hour, then this value is taken as the minimum water production constraint for that period. The upper limit of the temperature of the low-temperature multi-effect distillation process can be determined based on the tolerance range of the evaporator material, the empirical temperature of scaling, and long-term operation records. For example, the upper limit of control between 70°C and 85°C can be taken. The operating pressure range of reverse osmosis can be determined based on the rated parameters of the membrane module and the stable operating range. For example, it can be taken as 4.5MPa to 6.5MPa. The target range of feed ion concentration for each process can be determined by back-calculation based on the historical data of batches that meet the standards. For example, the distribution range of each ion concentration within 30 consecutive operating cycles that meet the standards can be statistically analyzed, and the middle 80% concentration range can be taken as the target window. In the equipment operation constraints, the upper and lower limits of each process load can be set according to 30% to 100% of the rated capacity, the state of charge of the energy storage device is preferably limited to 20% to 90%, and the rate of change of actuator action can be limited to no more than 10% of the rated value per cycle.

[0096] Furthermore, the aforementioned thresholds are not fixed but can be dynamically adjusted according to the operational stage. For example, in the initial stage of start-up, to avoid operational fluctuations, the load adjustment range can be appropriately narrowed, further reducing the single-cycle variation of the actuator from 10% to 5%; when the salinity of the raw material increases or the purity requirements of the target product increase, the target range of feed ion concentration in the extraction section can be appropriately tightened. Those skilled in the art will understand that the above-mentioned constraint values ​​can all be obtained from the equipment manual, commissioning data, and historical operating records. As long as they can reflect the permissible boundaries of the device and support subsequent solutions, they fall within the scope of this application.

[0097] S2.2.2: Within each control cycle, using the current global state vector as the initial condition, the future operating state of each process is predicted based on the dynamic mathematical model, and the optimized control instruction set is solved.

[0098] Specifically, the rolling forecasting method based on the control cycle is adopted because solar irradiance, time-of-use electricity prices, waste heat input, and the composition of concentrated brine all exhibit time-varying characteristics. If control values ​​are directly given based solely on current instantaneous data, they often only reflect the current operating conditions and cannot take into account changes in energy supply and material response lags in the next period. By using a dynamic mathematical model to extrapolate the state for several future periods in the current cycle, it is possible to predict in advance the available energy capacity, the impact of desalination section load changes on concentrated brine output, and the response of the extraction section to changes in feed. This allows control commands to no longer rely on single-point judgments but to be based on short-term trend forecasting.

[0099] In this embodiment, the control period is preferably set to 5 to 15 minutes, and the prediction time domain preferably covers the subsequent 3 to 6 control periods. Each time a solution is obtained, information such as solar thermal power, electricity price, energy storage state of charge, distillation temperature, reverse osmosis pressure, brine flow rate and concentration, reagent dosage for each process, and product purity from the current global state vector is first read. This information is then substituted into the energy supply sub-model, seawater desalination sub-model, and chemical resource extraction sub-model to predict the remaining thermal energy, electricity demand, water production capacity, composition of the mixed brine, and the potential yield levels for each extraction process in future periods. Subsequently, using the target load of the low-temperature multi-effect distillation process, the target load of the reverse osmosis process, the target power of the microwave heat source, the charging and discharging power of the energy storage device, and the reagent dosage for each process as candidate control variables, candidate combinations are searched under the aforementioned constraints and ranked according to the comprehensive objective comprised of energy consumption cost and comprehensive resource extraction benefits. For example, in the water supply priority scenario, the energy consumption cost weight can be set to 0.45, the resource extraction revenue weight to 0.35, and the operation stability correction weight to 0.20; in the resource recycling priority scenario, these can be adjusted to 0.35, 0.45, and 0.20.

[0100] Furthermore, to avoid frequent jumps in the solution results, a penalty for changes between adjacent control cycles is preferentially added during the sorting process. That is, if a candidate combination has a lower cost, but its distillation load, reverse osmosis load, or reagent dosage changes too much compared to the previous cycle, its priority is reduced. After this processing, the resulting optimized control instruction set reflects both the short-term future operating trend and maintains continuity with the current operating conditions, which can reduce problems such as thermal inertia mismatch, membrane segment shock, and excessive reagent dosage caused by abrupt adjustments.

[0101] S2.2.3: Execute the control instruction corresponding to the current control cycle in the optimized control instruction set, and re-solve the problem in the next control cycle with the updated global state vector as the initial condition to achieve rolling optimization.

[0102] Specifically, during the rolling optimization process, not all results within the entire prediction time domain are executed directly, but only the control quantities corresponding to the current control cycle are executed. This is because solar irradiance, electricity price, waste heat access, and process status may still change in the next cycle. If all control results from multiple future cycles were fixed and distributed at once, the adaptability to new operating conditions would be weakened. By retaining the processing logic of "predicting the future, executing the present, and recalculating in the next cycle," subsequent decisions can be continuously revised using the latest measured data while taking into account trend judgments.

[0103] In this embodiment, the control command corresponding to the current control cycle can be specifically manifested as follows: issuing an evaporation load adjustment value to the low-temperature multi-effect distillation actuator, issuing a target operating load or target pressure correction value to the reverse osmosis actuator, issuing a target power value to the microwave heat source, issuing a charging or discharging mode and corresponding power value to the energy storage device, and issuing adjustment values ​​such as feed rate, evaporation load, oxidant dosage, precipitant dosage, adsorbent dosage, or elution frequency to the evaporation crystallization, bromine extraction, magnesium / potassium extraction, and lithium extraction processes, respectively. After the control command is issued, it is preferable to wait for a sampling window to obtain execution feedback. The feedback includes at least the actual position value of the actuator, the current water production, the current concentrated brine concentration, the main ion concentration, and the product purity. If the deviation between the actual execution value and the issued value exceeds a preset threshold, the deviation is written into the updated global state vector. This deviation threshold can be determined according to the accuracy of the actuator. For example, load-type actuators can be set to ±3% of the target value, and reagent dosing actuators can be set to ±5% of the target value.

[0104] Furthermore, upon entering the next control cycle, the prediction results from the previous round are no longer used. Instead, the global state vector updated by sampling feedback is used to re-enter the prediction and solution process in S2.2.2. This approach ensures that control deviations not fully executed in the previous cycle, new changes in the external environment, and quality feedback from the extraction stage can all be incorporated into the next optimization cycle in real time, keeping the optimization process focused on the actual state of the equipment. Consequently, the adjustment relationships between low-temperature multi-effect distillation, reverse osmosis, and subsequent extraction processes can be continuously corrected and adjusted, preventing control failure due to a gradual disconnect between model predictions and on-site execution.

[0105] The specific steps for S3 are as follows:

[0106] S3.1: Based on the solar irradiance prediction sequence and the time-of-use electricity price sequence, determine the current operating scenario and switch between high heat energy supply scenario, low heat energy off-peak electricity price scenario and low heat energy peak electricity price scenario, dynamically adjusting the load distribution of the low temperature multi-effect distillation process and the reverse osmosis process.

[0107] Specifically, the operational scenario should not be determined solely based on solar energy or electricity prices at a single moment. Instead, it needs to be comprehensively determined by considering the predicted solar irradiance for the next control cycle, the current thermal storage status, the time-of-use electricity price period, and the heat demand corresponding to the current target load of low-temperature multi-effect distillation. This approach is necessary because low-temperature multi-effect distillation is heat-driven, while reverse osmosis is electricity-driven. The economics of both change synchronously with the availability of thermal energy and the electricity price level. If the scenario is not divided first, subsequent load adjustments are prone to resulting in wasted thermal energy or increased electricity consumption.

[0108] In this embodiment, the solar thermal power can be obtained by combining the irradiance prediction sequence with the collector area, collector efficiency, and current heat exchange loss correction value, while the time-of-use electricity price is directly read according to the scheduling period. The peak electricity price threshold is preferably determined by the average of the peak electricity prices of the past 30 operating days, and can be exemplarily set at 0.85 yuan / kWh; the waste heat access threshold can be set at 60% to 70% of the minimum stable heat load of low-temperature multi-effect distillation, and can be exemplarily set at 65% of the design evaporation heat load. After completing the scenario determination, the candidate load range obtained in S2 is then called for allocation.

[0109] Furthermore, scene switching employs a cycle-by-cycle re-judgment method instead of fixed-time-period switching. If the judgment results of two consecutive cycles differ, the prediction result of the later cycle is preferred, while limiting the single-cycle variation of distillation load and reverse osmosis load, for example, controlling it within 10% of the rated load. This avoids frequent switching caused by short-term fluctuations in solar irradiance and maintains stable load distribution.

[0110] For example, in a high heat energy supply scenario, when the solar thermal power is not lower than the rated heat load of the low-temperature multi-effect distillation process, the operating load of the low-temperature multi-effect distillation process is increased to prioritize the absorption of solar thermal power, the operating load of the reverse osmosis process is reduced accordingly, and the excess heat energy is stored in the thermal storage device.

[0111] When the solar thermal power is not lower than the rated heat load of the low-temperature multi-effect distillation process, it means that the heat side has the conditions to support the efficient operation of the distillation section. If a high reverse osmosis load is maintained at this time, the available thermal energy will be idle and the overall water production cost will not be reduced sufficiently. Therefore, it is necessary to prioritize increasing the operating load of the low-temperature multi-effect distillation process.

[0112] In this embodiment, the low-temperature multi-effect distillation load can be first increased to 80% to 100% of the rated load, and then the reverse osmosis load can be reduced in the reverse direction according to the total water production demand. If there is still surplus solar thermal power, the remaining portion can be introduced into a thermal storage device. The thermal storage device can be a hot water storage tank or a phase change thermal storage unit, and the upper limit of thermal storage is preferably determined according to the distillation heat replenishment demand for the subsequent 2 to 3 control cycles. This process can convert short-term surplus heat into an adjustable heat source for the subsequent period.

[0113] Furthermore, if the thermal storage device is near full load, it is preferable to maintain the low-temperature multi-effect distillation at a stable operating level near its rated value and continue to reduce the reverse osmosis power to avoid heat waste due to insufficient heat absorption. In this scenario, after load adjustment, the flow rate of concentrated brine entering the subsequent extraction stage usually increases, and the concentration degree becomes more stable, which is beneficial to maintaining continuous subsequent reagent conditioning.

[0114] For example, in a low-heat-energy, low-off-peak-price scenario, when the solar thermal power is lower than the heat demand corresponding to the current target load of the low-temperature multi-effect distillation process, and the time-of-use electricity price is not higher than the preset peak electricity price threshold, the reverse osmosis process load is increased to utilize low-priced electricity to produce water, while the microwave heat source is activated to supplement the heat required by the low-temperature multi-effect distillation process to maintain its minimum operating load.

[0115] When the solar thermal power is lower than the heat demand corresponding to the current target load of low-temperature multi-effect distillation, and the electricity price is at a low point, continuing to maintain a high distillation load will bring additional heating costs. In this case, shifting the focus of water production to reverse osmosis is more in line with energy utilization characteristics. Therefore, the method of "increasing the load through reverse osmosis and ensuring the bottom operation through distillation" is adopted.

[0116] In this embodiment, the reverse osmosis load can preferably be increased to 70% to 95% of the rated load, while the low-temperature multi-effect distillation is maintained at the minimum stable operating load, which can be 35% to 45% of the rated load, for example. In this scenario, the microwave heat source only serves to supplement the minimum heat demand of the distillation section; its power is not at its rated value, but is determined by the difference between the minimum stable heat load and the current amount of heat that can be provided by solar energy. This method avoids the thermal inertia loss caused by completely shutting down and restarting the distillation section.

[0117] Furthermore, if the off-peak electricity price lasts longer than two control cycles, the charging priority of the energy storage device can be appropriately increased. After meeting the reverse osmosis load, the remaining low-priced electricity can be used for supplementary energy supply, so that it can be released during the subsequent peak electricity price period. This approach can both utilize low-priced electricity to expand water production capacity and maintain the continuous operation of the distillation section.

[0118] For example, in a scenario with low peak electricity price for heat energy, when the solar thermal power is lower than the heat demand corresponding to the current target load of the low-temperature multi-effect distillation process, and the time-of-use electricity price is higher than the preset peak electricity price threshold, the low-temperature multi-effect distillation process and the reverse osmosis process are simultaneously reduced to the minimum operating load, and the energy storage device is called to release electrical energy to supplement the electricity required for basic water production.

[0119] When the solar thermal power is insufficient and the time-of-use electricity price enters its peak, maintaining a high reverse osmosis load or high-power supplemental heating will directly increase the unit water production cost. Therefore, in this scenario, it is necessary to reduce the distillation section and reverse osmosis section to the minimum operating load at the same time to maintain basic water supply and avoid high-cost expansion.

[0120] In this embodiment, the minimum operating load for both low-temperature multi-effect distillation and reverse osmosis can be determined as 30% to 40% of their respective rated loads. Specific values ​​can be obtained by filtering historical stable operating data, i.e., selecting the minimum load that allows for continuous operation for at least 2 hours and ensures the product water quality meets standards as the benchmark. In this scenario, the energy storage device enters a discharge-priority state. The discharge power is preferably determined by subtracting the current allowable power purchase from the basic water production electricity demand, but the state of charge must not fall below 20%.

[0121] Furthermore, if the peak electricity price is predicted to last only one control cycle, it is preferable to use energy storage to supplement power and maintain basic operation; if the peak is predicted to last for multiple cycles, the load will be reduced to a lower level in the first cycle and the energy storage release time will be extended. Through this approach, energy consumption during peak electricity price periods is controlled within the range of basic water supply.

[0122] Furthermore, when the power plant waste heat supply reaches the preset waste heat access threshold, the power plant waste heat is identified as the priority heat source for the low-temperature multi-effect distillation process. The output of solar thermal collectors is reduced and the output of microwave heat source is lowered. Based on the power plant waste heat, the target load of the low-temperature multi-effect distillation process is recalculated, and the target load of the reverse osmosis process is reduced accordingly.

[0123] Specifically, once the waste heat reaches a level that can be stably utilized, its heat quality and heating continuity are usually better than short-term fluctuations in solar supplementary heating. Therefore, after meeting the preset access threshold, it is preferentially allocated to the low-temperature multi-effect distillation section, which can reduce the use of microwave heat source and reconstruct the load ratio between distillation and reverse osmosis.

[0124] In this embodiment, the waste heat supply is preferably determined based on the average available heat flow over two consecutive control cycles to avoid accidental triggering due to instantaneous fluctuations. Once the threshold is reached, the minimum waste heat share required for the target operation of the low-temperature multi-effect distillation is locked first, and then the solar collector output allocation and microwave heat source output are adjusted accordingly. If necessary, only the microwave heat source is retained as a dynamic compensation boundary. The recalculated distillation target load can be increased by 10% to 25% compared to the original value, while the reverse osmosis target load is adjusted downwards to balance the total permeable water demand.

[0125] Furthermore, if the waste heat is continuously supplied for more than a preset time, such as 30 minutes, the priority of the distillation section in the subsequent two prediction cycles can be increased simultaneously. If the waste heat is interrupted, it will be reassigned according to the scenario of the next control cycle, without using the original waste heat priority result. This process is beneficial for directly converting an external, inexpensive heat source into a stable source of concentrated brine.

[0126] S3.2: Based on the dynamic mathematical model, predict the change in the concentration of concentrated brine in the next control cycle, and feed the change in the concentration of concentrated brine and the predicted concentration of each ion forward to the control unit of each process of chemical resource extraction, so as to control each process to dynamically adjust the dosage of reagents according to the concentration of feed ions, forming a cascade parameter linkage control.

[0127] For example, the cascade parameter control process includes:

[0128] Based on the change in the concentration of the concentrated brine, the feed rate and evaporation load of the evaporation and crystallization process are adjusted, and the predicted concentrations of each ion in the output mother liquor are sequentially fed forward to the bromine extraction, magnesium / potassium extraction and lithium extraction processes.

[0129] The bromine extraction process dynamically adjusts the oxidant dosage based on the bromide ion concentration in the feed, and transfers the ion concentration in the mother liquor after bromine extraction to the magnesium / potassium extraction process.

[0130] The magnesium / potassium extraction process dynamically adjusts the amount of precipitant added based on the magnesium and potassium ion concentrations in the feed, and potassium extraction is carried out after the magnesium removal meets the standard. The ion concentration in the mother liquor after magnesium / potassium extraction is then transferred to the lithium extraction process.

[0131] The lithium extraction process dynamically adjusts the amount of adsorbent and the elution frequency based on the lithium ion concentration and magnesium-lithium ratio in the feed.

[0132] Specifically, after the load distribution in the desalination section changes, the first thing affected is the flow rate, total concentration, and main ionic composition of the concentrated brine entering the extraction section. If the extraction section continues to use the fixed dosage of reagents from the previous cycle, local over-dosing, insufficient reaction, or separation window shift may easily occur. Therefore, feeding forward the changes in the concentration of concentrated brine and the predicted concentrations of each ion to each process is to adjust the parameters before the material actually enters the process.

[0133] In this embodiment, the evaporation crystallization process adjusts the feed rate and evaporation load based on changes in the concentrated brine concentration. When the predicted concentration increases, the feed rate is appropriately reduced and the evaporation load is increased; when the predicted concentration decreases, the feed rate is appropriately restored to avoid over-evaporation. The concentrations of the main ions in the output mother liquor are then sequentially transferred to the bromine extraction, magnesium / potassium extraction, and lithium extraction processes. The bromine extraction process preferably determines the oxidant addition range based on the predicted bromide ion concentration. The magnesium / potassium extraction process adjusts the precipitant addition sequence and amount based on the magnesium and potassium ion concentrations, respectively. The lithium extraction process simultaneously considers the lithium ion concentration and the magnesium-lithium ratio to determine the adsorbent dosage and elution frequency.

[0134] Furthermore, the adjustment of reagents in each process does not adopt a one-step jump method, but is adjusted based on the dosage value of the previous cycle, with the adjustment range in a single cycle preferably controlled between 5% and 12% of the original dosage value. The magnesium-lithium ratio threshold can be determined based on the statistics of historical batches that meet the standards. For example, the average magnesium-lithium ratio in 20 consecutive batches of lithium products that meet the purity standards can be used as a benchmark, and a range of 10% above and below it can be used as an adaptation window.

[0135] The specific steps for S4 are as follows:

[0136] S4.1: When the purity of the product in any process is lower than the set purity threshold, detect whether the feed ion concentration of that process deviates from the preset normal concentration range.

[0137] Specifically, in this embodiment, when the product purity of a certain process is lower than a preset purity threshold, the first step is to determine whether there is an abnormality in the ion concentration of the feed. This step is to ensure that fluctuations in product purity are not solely caused by process fluctuations, but rather by fluctuations in the quality of upstream feed materials or changes in input conditions. By detecting whether the feed ion concentration deviates from the preset normal concentration range, purity abnormalities caused by changes in the feed can be detected in a timely manner, avoiding errors that cannot be corrected by simple process adjustments. This step, by setting a monitoring range for ion concentration, identifies potential raw material problems early on, ensuring that the extraction process can be adjusted in a timely manner, and preventing the amplification of purity deviations in subsequent processes.

[0138] In this embodiment, the normal range of feed ion concentration can be determined based on historical operating data, equipment performance indicators, and product quality standards. For example, the feed concentration after seawater desalination can be set between 1000-2000 mg / L for sodium ions and between 200-400 mg / L for magnesium ions. These values ​​are based on common values ​​observed during previous stable operation. If the feed concentration deviates from the normal range, it indicates a potential raw material problem that requires feedback and correction.

[0139] S4.2: If the feed ion concentration of this process does not deviate from the preset normal concentration range, it is determined that the process of this process is abnormal, and the dosage of the reagent in this process is adjusted.

[0140] Specifically, when the feed concentration is within the normal range, the problem is likely due to an abnormality in the process operation. In this case, the approach of this embodiment is to adjust the dosage of the reagent in this process to correct the abnormality in product purity. The adjustment of the reagent dosage depends on historical operating experience and the current process status. By dynamically adjusting the amount of reagent added, the decrease in product purity caused by insufficient reaction rate or low ion removal efficiency is compensated.

[0141] In this embodiment, the adjustment of reagent dosage is not set by a single rule, but rather dynamically corrected by combining historical data and real-time feedback. For example, in the evaporation and crystallization process, if the product purity is low, it may be due to incomplete removal of bromide ions. In this case, the dosage of oxidant will be increased to accelerate the reaction. In the lithium extraction process, it may be due to a decrease in lithium ion adsorption rate, so the amount of adsorbent will be reduced and the elution frequency will be extended. By precisely adjusting the reagents in each process, the decrease in product purity caused by process fluctuations can be minimized.

[0142] Furthermore, the adjustment range for reagents is typically set between 5% and 15% of the current dosage to avoid secondary pollution or resource waste caused by overdosing. For each process's reagent adjustment, historical data is compared and analyzed to ensure that every adjustment is based on accurate process feedback.

[0143] S4.3: If the feed ion concentration of this process deviates from the preset normal concentration range, it is determined that the raw material quality is abnormal, and an abnormal signal is fed back to the model prediction control algorithm.

[0144] Specifically, in this embodiment, the determination of abnormal raw material quality is based on deviations in the feed ion concentration. If the feed concentration exceeds the normal range, it indicates a potential external quality problem, such as abnormal seawater salinity or improper feed pretreatment. In this case, the abnormal signal is immediately fed back to the model predictive control algorithm to trigger adjustments in the front-end seawater desalination process, further correcting the input parameters in the desalination process.

[0145] In this embodiment, the abnormal signal feedback process includes the linkage between real-time data acquisition and front-end process correction. When the feed ion concentration exceeds the set range, for example, sodium ion concentration exceeds 2000 mg / L, or magnesium ion concentration is below 200 mg / L, the control system automatically sends a feedback signal to increase the concentration ratio of the desalination section and reduce the dissolved salinity, thereby improving the ion removal efficiency of the subsequent extraction section. The feedback mechanism can adjust process parameters in a timely manner according to changes in feed, reducing the impact of unstable raw material quality.

[0146] Furthermore, abnormal signals in the feed quality can be automatically identified by the integrated monitoring module in the system and transmitted in real time to upstream operators or the scheduling system via remote sensors or controllers, thereby achieving rapid response. This mechanism ensures rapid feedback and adjustment to fluctuations in raw material quality, reducing production instability caused by changes in raw materials.

[0147] S4.4: The model predictive control algorithm improves the concentration of the brine output from the seawater desalination process based on the abnormal signal, so as to increase the concentration of the target ions to be extracted.

[0148] Specifically, when the system receives an abnormal signal from the chemical resource extraction process, the model predictive control algorithm will adjust the seawater desalination process based on the nature of the abnormal signal. This adjustment primarily affects the feed conditions of the extraction process by adjusting the concentration of the brine. The adjustment of the brine concentration directly relates to the concentration level of target ions in subsequent extraction processes.

[0149] In this embodiment, the model predictive control algorithm dynamically adjusts the load of the low-temperature multi-effect distillation and reverse osmosis processes based on real-time feedback of changes in the concentrated brine concentration, ensuring that the ion concentration of the output concentrated brine meets the requirements of the extraction process. If the extraction process requires increasing the concentration of resources such as magnesium, potassium, or lithium, the control system increases the concentration ratio of the evaporation crystallization or reverse osmosis processes while reducing the water flow rate. In this way, the control algorithm can adjust the desalination process based on feedback signals, thereby improving the efficiency of subsequent extraction.

[0150] Furthermore, the concentration of the brine is not adjusted solely based on static setpoints, but rather dynamically calculated according to the deviation between real-time operating data, feedback signals, and the target purity. This ensures stable operation of subsequent extraction processes and efficient resource recovery.

[0151] S4.5: Feed the corrected brine concentration prediction value forward to the control unit of each chemical resource extraction process to trigger the readjustment of reagent dosage in each process, thereby forming a corrective closed loop.

[0152] Specifically, the revised predicted concentration of the brine will be fed back to the chemical resource extraction process as feedback, used to adjust the dosage of reagents in subsequent steps. This operation ensures that the resource extraction process not only responds to the current state of the brine but also adjusts in real time according to changes in the upstream process. Through this closed-loop control mechanism, subsequent processes can respond promptly based on the latest material conditions, rather than simply relying on statically set reagent dosages from the raw material input.

[0153] In this embodiment, the corrected predicted brine concentration is fed forward to the control units of extraction processes such as bromine, magnesium / potassium, and lithium extraction via an embedded transmission module. The specific reagent dosage is adjusted in real time based on the deviation between the predicted concentration and the target purity. For example, if the predicted bromide ion concentration increases, the bromine extraction process will automatically increase the oxidant dosage to ensure bromine removal efficiency; if the lithium ion concentration is insufficient, the amount of adsorbent and the elution frequency will be adjusted to supplement lithium recovery.

[0154] Furthermore, the implementation of feedforward correction ensures rapid response in subsequent processes without relying on offline analysis or periodic adjustments. This technical solution enables each step to be operated not merely as an adjustment to the real-time state, but as a proactive adjustment based on the future state, greatly improving the stability of the extraction process and resource recovery rate.

[0155] The specific steps for S5 are as follows:

[0156] S5.1: Count the number of control cycles in which the deviation exceeds the set threshold consecutively. When the number of control cycles reaches the preset consecutive cycle number threshold, trigger online identification and update.

[0157] Specifically, when the deviation between the actual operating results and the model prediction results within a control cycle exceeds a set threshold, it indicates a significant discrepancy between the system's operating state and the predictive model. This could be due to changes in the external environment, equipment aging, or changes in process parameters. To ensure the continuous effectiveness of the control system, it is necessary to determine whether to trigger online identification and updates by statistically analyzing the consecutive occurrences of deviations. When deviations exceed the threshold multiple times consecutively, it indicates that the model may no longer be suitable for the current operating conditions. In this case, online identification and updates are required to correct the model's parameters, ensuring that the model can reflect the actual operating state in real time.

[0158] In this embodiment, the set deviation threshold is determined based on long-term operating data and historical error distribution. For example, assuming the deviation threshold for the comprehensive benefit index is set at 5%, online identification and updating are triggered when the deviation exceeds this threshold for three consecutive periods. This threshold and triggering condition can be adjusted appropriately according to actual operating conditions and equipment stability. This triggering method based on deviation continuity not only avoids erroneous updates caused by single errors but also ensures that the system can respond to changes promptly, preventing long-term accumulation of errors from affecting control performance.

[0159] S5.2: Collect actual operating data within a preset quantity control period before the triggering time, and identify and update the key parameters of the energy supply sub-model, the seawater desalination sub-model, and the chemical resource extraction sub-model respectively.

[0160] Specifically, the core of online identification and updating lies in adjusting the key parameters of the model based on actual operating data. After an update is triggered, real-time data from several past control cycles is collected to analyze the causes of deviations, and the parameters of each sub-model are optimized based on the actual operating status. In this embodiment, the "key parameters" include, but are not limited to, the thermal and electrical supply capacity in the energy supply sub-model, the concentration ratio in the seawater desalination sub-model, the recovery rate of reverse osmosis, and the response relationship between reagent dosage and yield in the extraction sub-model.

[0161] In this embodiment, the collected data includes key information such as real-time power, temperature, pressure, flow rate, and concentration within each control cycle, with a data acquisition cycle ranging from 5 to 30 minutes. By comparing these actual operating data with historical data and considering model prediction deviations, parameters requiring updates are identified and optimized using the least squares method or weighted average method. This adjustment process ensures that the model parameters can promptly match actual operating conditions when external conditions or equipment status change, thereby guaranteeing the accuracy of subsequent predictions.

[0162] Furthermore, by periodically analyzing and reviewing historical data, the update frequency can be optimized to avoid system oscillations caused by over-adjustment. For example, when the equipment enters a stable operating phase, the frequency of online updates can be reduced; while during startup or periods of significant load change, the update frequency should be increased. This dynamic online adjustment mechanism helps ensure that the system maintains high control accuracy under various operating conditions.

[0163] S5.3: Write the updated model parameters into the model predictive control algorithm, and it will take effect from the next control cycle.

[0164] Specifically, once the model parameters are updated, these parameters must be promptly rewritten into the model predictive control algorithm so that at the start of the next control cycle, the new control commands can be optimized based on the latest model state. The key to this step is ensuring that the updated model can respond to actual operating conditions in the shortest possible time and that the optimized control algorithm can perform rolling predictions and adjustments based on the new parameters, thereby guaranteeing the accuracy of subsequent control.

[0165] In this embodiment, the model parameter updates include not only static parameters, such as thermal energy supply efficiency and membrane module recovery rate, but also dynamic parameters, such as the response relationship between feed concentration and reagent dosage. The updated parameters are loaded into the control algorithm at the beginning of the control cycle and used to solve for optimized control commands in subsequent cycles. This process is completed through an automated interface, reducing the complexity of manual intervention and system adjustments.

[0166] Furthermore, the updated parameters not only take effect within the current control cycle but also influence optimization decisions for multiple future control cycles. This approach ensures that the model can respond quickly and accurately to all influencing factors under different operating conditions. For example, when solar irradiance changes significantly, online-updated model parameters can more precisely adjust the load distribution of low-temperature multi-effect distillation and reverse osmosis, thereby achieving optimal resource utilization efficiency and the lowest operating costs.

[0167] Example 2

[0168] Please see Figure 3 The present invention provides an embodiment of a comprehensive seawater resource utilization system based on multi-energy complementarity and intelligent regulation, the system comprising:

[0169] The data acquisition unit is used to collect operational status data and external environmental data of energy supply, seawater desalination and chemical resource extraction processes, wherein the external environmental data includes solar irradiance prediction sequences and time-of-use electricity price sequences.

[0170] A state construction unit is used to construct a global state vector based on the running state data and external environment data;

[0171] The dynamic modeling and optimization unit is used to construct dynamic mathematical models of energy supply, seawater desalination and chemical resource extraction processes based on the global state vector, and to solve the optimized control instruction set in a rolling manner.

[0172] The coordination and control unit is used to coordinate and control the load distribution of the seawater desalination process and the process parameters of the chemical resource extraction process according to the optimized control instruction set.

[0173] The anomaly correction unit is used to reverse the source and correct the operating parameters of the seawater desalination process when an abnormality in product purity occurs at any step in the chemical resource extraction process, forming a closed loop of correction.

[0174] The online update unit is used to identify and update the dynamic mathematical model online based on actual operating data when the deviation between the comprehensive benefit index obtained from actual operation and the comprehensive benefit index predicted by the dynamic mathematical model continues to exceed a set threshold.

[0175] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A comprehensive utilization method for seawater resources based on multi-energy complementarity and intelligent regulation, characterized in that, The method includes: The system collects operational status data and external environmental data of energy supply, seawater desalination and chemical resource extraction processes, and constructs a global state vector based on these data. The external environmental data includes solar irradiance prediction sequences and time-of-use electricity price sequences. Based on the global state vector, a dynamic mathematical model of the energy supply, seawater desalination and chemical resource extraction process is constructed, and the model predictive control algorithm is used to solve the optimization control instruction set in a rolling manner. According to the optimized control instruction set, the load allocation of the seawater desalination process and the process parameters of the chemical resource extraction process are coordinated and controlled. When any step in the chemical resource extraction process results in abnormal product purity, the process traces back to the source and corrects the operating parameters of the seawater desalination process, forming a corrective closed loop. When the deviation between the comprehensive benefit index obtained from actual operation and the comprehensive benefit index predicted by the dynamic mathematical model continues to exceed a set threshold, the dynamic mathematical model is identified and updated online based on the actual operation data to achieve the synergistic and efficient utilization of water resources and chemical resources in seawater.

2. The method for comprehensive utilization of seawater resources based on multi-energy complementarity and intelligent regulation according to claim 1, characterized in that, The construction of the dynamic mathematical model and the rolling solution optimization of the control instruction set include: Based on the global state vector, an energy supply sub-model, a seawater desalination sub-model, and a chemical resource extraction sub-model are constructed respectively to form the dynamic mathematical model. Under the premise of satisfying the process constraints and equipment operation constraints of each process, the comprehensive objective function consisting of energy consumption cost and comprehensive resource extraction benefits is used as the optimization objective, and the optimization control instruction set is solved in a rolling manner. The optimization control instruction set includes the target load of low temperature multi-effect distillation process, the target load of reverse osmosis process, the target power of microwave heat source, the charging and discharging power of energy storage device, and the amount of reagents added in each process of chemical resource extraction.

3. The method for comprehensive utilization of seawater resources based on multi-energy complementarity and intelligent regulation according to claim 2, characterized in that, The energy supply sub-model, seawater desalination sub-model, and chemical resource extraction sub-model respectively include: The energy supply sub-model uses the energy supply parameters of solar thermal collectors, grid power supply, microwave heat source, energy storage device charging and discharging parameters, and power plant waste heat as inputs to establish dynamic energy balance equations for the energy load of each process flow. The seawater desalination sub-model uses the heat source parameters of the low-temperature multi-effect distillation process and the operating pressure of the reverse osmosis process as inputs to calculate the flow rate and concentration of the concentrated brine output from the low-temperature multi-effect distillation process and the reverse osmosis process, respectively, and calculates the concentration of the concentrated brine after mixing accordingly. The chemical resource extraction sub-model takes the feed ion concentration and reagent dosage of the evaporation crystallization, bromine extraction, magnesium / potassium extraction and lithium extraction processes as inputs, calculates the product yield of each process, and sequentially transfers the ion concentration in the output mother liquor of each process to the next process as the feed ion concentration.

4. The method for comprehensive utilization of seawater resources based on multi-energy complementarity and intelligent regulation according to claim 2, characterized in that, The rolling solution optimization control instruction set includes: The process constraints and equipment operation constraints of each process flow are used as the solution conditions. The process constraints include the water production rate not being lower than the preset demand, the upper limit of the temperature of the low temperature multi-effect distillation process, the operating pressure range of the reverse osmosis process, and the target range of feed ion concentration for each process. The equipment operation constraints include the upper and lower limits of the load of each process flow, the state of charge range of the energy storage device, and the rate of change of action of each actuator. Within each control cycle, using the current global state vector as the initial condition, the future operating state of each process is predicted based on the dynamic mathematical model, and the optimized control instruction set is solved. The control instructions corresponding to the current control cycle are executed from the optimized control instruction set, and the solution is recalculated in the next control cycle with the updated global state vector as the initial condition to achieve rolling optimization.

5. The method for comprehensive utilization of seawater resources based on multi-energy complementarity and intelligent regulation according to claim 1, characterized in that, The coordination control includes: Based on the solar irradiance prediction sequence and the time-of-use electricity price sequence, the current operating scenario is determined, and the system switches between high heat energy supply scenario, low heat energy off-peak electricity price scenario and low heat energy peak electricity price scenario to dynamically adjust the load distribution of the low temperature multi-effect distillation process and the reverse osmosis process. The change in brine concentration in the next control cycle is predicted based on the dynamic mathematical model. The change in brine concentration and the predicted concentration of each ion are then fed forward to the control unit of each chemical resource extraction process to control the dynamic adjustment of reagent dosage according to the feed ion concentration, thus forming a cascade parameter linkage control.

6. The method for comprehensive utilization of seawater resources based on multi-energy complementarity and intelligent regulation according to claim 5, characterized in that, The dynamic adjustment of the load distribution between the low-temperature multi-effect distillation process and the reverse osmosis process includes: In high heat energy supply scenarios, when the solar thermal power is not lower than the rated heat load of the low temperature multi-effect distillation process, the operating load of the low temperature multi-effect distillation process is increased to prioritize the consumption of solar thermal power, the operating load of the reverse osmosis process is reduced accordingly, and the excess heat energy is stored in the thermal storage device. In scenarios with low heat energy and low off-peak electricity prices, when the solar thermal power is lower than the heat demand corresponding to the current target load of the low-temperature multi-effect distillation process, and the time-of-use electricity price is not higher than the preset peak electricity price threshold, the reverse osmosis process load is increased to use low-priced electricity to produce water, and at the same time, the microwave heat source is activated to supplement the heat required by the low-temperature multi-effect distillation process to maintain its minimum operating load. In scenarios with low heat energy and peak electricity price, when the solar thermal power is lower than the heat demand corresponding to the current target load of the low temperature multi-effect distillation process, and the time-of-use electricity price is higher than the preset peak electricity price threshold, the low temperature multi-effect distillation process and the reverse osmosis process are simultaneously reduced to the minimum operating load, and the energy storage device is called to release electrical energy to supplement the electricity required for basic water production. When the waste heat supply from the power plant reaches the preset waste heat access threshold, the waste heat from the power plant is identified as the priority heat source for the low-temperature multi-effect distillation process. The output of the solar thermal collector is reduced and the output of the microwave heat source is lowered. Based on the waste heat from the power plant, the target load of the low-temperature multi-effect distillation process is recalculated, and the target load of the reverse osmosis process is reduced accordingly.

7. The method for comprehensive utilization of seawater resources based on multi-energy complementarity and intelligent regulation according to claim 5, characterized in that, The cascaded parameter linkage control includes: Based on the change in the concentration of the concentrated brine, the feed rate and evaporation load of the evaporation and crystallization process are adjusted, and the predicted concentrations of each ion in the output mother liquor are sequentially fed forward to the bromine extraction, magnesium / potassium extraction and lithium extraction processes. The bromine extraction process dynamically adjusts the oxidant dosage based on the bromide ion concentration in the feed, and transfers the ion concentration in the mother liquor after bromine extraction to the magnesium / potassium extraction process. The magnesium / potassium extraction process dynamically adjusts the amount of precipitant added based on the magnesium and potassium ion concentrations in the feed, and potassium extraction is carried out after the magnesium removal meets the standard. The ion concentration in the mother liquor after magnesium / potassium extraction is then transferred to the lithium extraction process. The lithium extraction process dynamically adjusts the amount of adsorbent and the elution frequency based on the lithium ion concentration and magnesium-lithium ratio in the feed.

8. The method for comprehensive utilization of seawater resources based on multi-energy complementarity and intelligent regulation according to claim 1, characterized in that, The process of reverse tracing and correcting the operating parameters of the seawater desalination process includes: When the purity of the product in any process is lower than the set purity threshold, the feed ion concentration of that process is checked to see if it deviates from the preset normal concentration range. If the feed ion concentration of this process does not deviate from the preset normal concentration range, it is determined that the process is abnormal and the dosage of the reagent in this process is adjusted. If the feed ion concentration of this process deviates from the preset normal concentration range, it is determined that the raw material quality is abnormal, and an abnormal signal is fed back to the model prediction control algorithm. The model predictive control algorithm increases the concentration of the brine output from the seawater desalination process based on the abnormal signal, thereby increasing the concentration of the target ions to be extracted. The corrected brine concentration prediction is fed forward to the control units of each chemical resource extraction process to trigger a readjustment of the reagent dosage in each process, thus forming a corrective closed loop.

9. The method for comprehensive utilization of seawater resources based on multi-energy complementarity and intelligent regulation according to claim 2, characterized in that, The online identification and updating of the dynamic mathematical model based on actual operational data includes: The number of control cycles in which the deviation exceeds a set threshold is counted. When the number of control cycles reaches a preset threshold for consecutive cycles, an online identification update is triggered. Collect actual operating data within a preset quantity control period before the triggering time, and identify and update the key parameters of the energy supply sub-model, the seawater desalination sub-model, and the chemical resource extraction sub-model respectively; The updated model parameters are written into the model predictive control algorithm and take effect from the next control cycle.

10. A comprehensive seawater resource utilization system based on multi-energy complementarity and intelligent regulation, used to implement the comprehensive seawater resource utilization method based on multi-energy complementarity and intelligent regulation as described in any one of claims 1-9, characterized in that, The system includes: The data acquisition unit is used to collect operational status data and external environmental data of energy supply, seawater desalination and chemical resource extraction processes, wherein the external environmental data includes solar irradiance prediction sequences and time-of-use electricity price sequences. A state construction unit is used to construct a global state vector based on the running state data and external environment data; The dynamic modeling and optimization unit is used to construct dynamic mathematical models of energy supply, seawater desalination and chemical resource extraction processes based on the global state vector, and to solve the optimized control instruction set in a rolling manner. The coordination and control unit is used to coordinate and control the load distribution of the seawater desalination process and the process parameters of the chemical resource extraction process according to the optimized control instruction set. The anomaly correction unit is used to reverse the source and correct the operating parameters of the seawater desalination process when an abnormality in product purity occurs at any step in the chemical resource extraction process, forming a closed loop of correction. The online update unit is used to identify and update the dynamic mathematical model online based on actual operating data when the deviation between the comprehensive benefit index obtained from actual operation and the comprehensive benefit index predicted by the dynamic mathematical model continues to exceed a set threshold.

Images