Ultralow-temperature MVR (mechanical vapor recompression) coupled reverse Carnot cycle zero-emission treatment method for methylnitroguanidine wastewater
By combining ultra-low temperature MVR evaporation and concentration with reverse Carnot cycle DTB crystallization and an intelligent adjustment system, the problems of high energy consumption and unstable operation in the treatment of methylnitroguanidine wastewater have been solved, achieving zero discharge and resource utilization of wastewater and ensuring stable system operation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-02-02
- Publication Date
- 2026-03-31
AI Technical Summary
Existing technologies for treating methyl nitroguanidine wastewater suffer from problems such as high energy consumption, equipment scaling, unstable operation, and lack of intelligent control, making it difficult to achieve complete purification and resource utilization of high-salt and high-organic wastewater.
The process employs a coupled ultra-low temperature MVR evaporation concentration and reverse Carnot cycle DTB crystallization, combined with an intelligent adjustment system. Through real-time data acquisition and multi-timescale predictive control, the steam compressor speed, forced circulation pump flow rate, and feed rate are automatically adjusted to achieve zero wastewater discharge and resource utilization.
It achieves complete resource utilization and zero discharge of methylnitroguanidine wastewater, with stable and reliable treatment effect, stable system operation, and reduced energy consumption and equipment failure rate.
Smart Images
Figure CN121758028A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of methylnitroguanidine wastewater treatment technology, specifically a zero-discharge treatment method for methylnitroguanidine wastewater using ultra-low temperature MVR coupled with reverse Carnot cycle. Background Technology
[0002] Methylnitroguanidine is an important chemical intermediate widely used in pharmaceuticals, pesticides, dyes, and other fields. Its production process generates large quantities of high-concentration organic wastewater, characterized by high COD concentration, high total nitrogen content, high salinity (3-6%), and poor biodegradability, classifying it as recalcitrant industrial wastewater. Traditional treatment methods mainly include physicochemical methods, biological treatment methods, and evaporation crystallization methods; however, these methods have many problems in practical applications.
[0003] Physicochemical methods such as coagulation sedimentation, adsorption, and advanced oxidation can remove some organic matter, but they are costly, generate large amounts of sludge, and are difficult to completely purify wastewater. Biological treatment methods are poorly adapted to high-salt, high-concentration organic wastewater, microbial activity is inhibited, treatment effects are unstable, and effluent fails to meet discharge standards. While evaporation crystallization can recover salt, traditional multi-effect evaporators are energy-intensive, prone to scaling, and have high operating costs, and their effectiveness in removing organic matter is limited.
[0004] In recent years, mechanical vapor recompression (MVR) technology has been applied in wastewater treatment due to its energy-saving advantages. However, conventional MVR technology still suffers from problems such as severe scaling, unstable operation, and high energy consumption when treating high-salt and high-organic wastewater. At the same time, existing evaporation crystallization systems lack intelligent control methods and cannot automatically adjust operating parameters according to changes in influent load and water quality, resulting in low system efficiency, high energy consumption, and high equipment failure rate. Summary of the Invention
[0005] To address the shortcomings of existing technologies, this invention provides a zero-discharge treatment method for methylnitroguanidine wastewater using ultra-low temperature MVR coupled with reverse Carnot cycle, which solves the problems mentioned in the background section.
[0006] To achieve the above objectives, the present invention provides a method for zero-discharge treatment of methylnitroguanidine wastewater using ultra-low temperature MVR coupled with reverse Carnot cycle, comprising the following steps: Step 1: After the methyl nitroguanidine wastewater enters the equalization tank for homogenization and equalization, it passes through an oil removal device to remove floating oil, a multi-media filter to remove suspended solids, a stripping device to remove volatile organic compounds, and an ozone catalytic oxidation-aerated biological filter to degrade organic matter. Step 2: The pretreated wastewater is fed into the MVR evaporator for evaporation and concentration under ultra-low temperature conditions. The MVR evaporator compresses and heats the secondary steam and then uses it as a heat source for recycling. The concentrated liquid is then fed into a heated separator for further separation. Step 3: The concentrate from Step 2 is fed into the DTB crystallizer, where it undergoes low-temperature evaporation and crystallization under the action of the reverse Carnot cycle system. The refrigerant is compressed and heated by the compressor, releasing heat in the refrigerant condenser. After being depressurized by the expansion valve, it absorbs heat in the refrigerant evaporator. Step 4: After centrifugation, the crystalline product obtained in Step 3 is dried in a drying oven to obtain a solid salt product; Step 5: The intelligent adjustment system automatically adjusts the steam compressor speed, forced circulation pump flow rate, and feed rate according to load changes.
[0007] Preferably, step five: the method of automatically adjusting the steam compressor speed, forced circulation pump flow rate, and feed rate according to load changes through an intelligent adjustment system includes: S51: Establish a real-time data acquisition system to collect process parameters, equipment parameters, environmental parameters, and energy consumption parameters; perform data cleaning and feature extraction; S52: Construct load forecasting models, energy consumption optimization models, and fault diagnosis models; S53: Deploy the model on an edge computing gateway and use incremental learning algorithms to continuously optimize the model; S54: Employs multi-timescale predictive control to achieve wide-load adaptive adjustment and formulates optimal operation plans in conjunction with peak-valley electricity pricing; S55: Employ a combination of unsupervised and supervised learning methods for anomaly detection, predict the remaining lifespan of critical equipment, and build an equipment fault knowledge base. S56: Real-time monitoring of performance indicators, automatic adjustment of control parameters based on evaluation results, establishment of a data closed-loop mechanism for continuous optimization, real-time monitoring of control accuracy, response time, stability and energy consumption indicators, setting of evaluation thresholds, obtaining performance evaluation scores through performance evaluation, and finding the optimal combination of control parameters when the performance evaluation score is lower than the threshold, automatically adjusting the steam compressor speed, forced circulation pump flow rate and feed rate control parameters.
[0008] Preferably, step S55: employing a combination of unsupervised and supervised learning methods for anomaly detection, predicting the remaining lifespan of critical equipment, and constructing an equipment fault knowledge base, specifically including: S551: Collect equipment operation data, process parameters, environmental parameters and energy consumption data, preprocess the data using standardization or normalization methods, store the preprocessed data in a time series database, and establish a data tagging system, including normal state, abnormal state and fault type tags. S552: Extract time-domain features, frequency-domain features, and time-frequency-domain features from the raw data; S553: Density-based clustering algorithm is used to perform cluster analysis on normal operating condition data to establish normal operating condition data clusters. Anomaly samples are detected by the isolated forest algorithm, and the anomaly score of the samples is calculated. An autoencoder is used for unsupervised feature learning. Anomalies are detected by reconstruction error. The local anomaly factor algorithm is used to calculate the local density of the samples and identify local anomaly points. The detection results of multiple unsupervised detection methods are fused to obtain the final anomaly detection result. S554: Construct a fault diagnosis sample library, including normal samples and various fault samples. Fault types include mechanical faults, electrical faults, thermal faults, and control faults. Use machine learning algorithms such as support vector machines, random forests, and gradient boosting trees to establish a multi-class fault diagnosis model. S555: Establish an equipment degradation model, using a physical model-based approach, combining equipment operating parameters and physical laws to establish degradation equations, using a data-driven approach to extract equipment degradation features, using machine learning-based methods, including support vector regression, random forest regression, and neural network regression, to establish a remaining life prediction model, and using deep learning-based methods, including long short-term memory networks and gated recurrent units, to establish a sequence prediction model to predict the remaining life of the equipment.
[0009] Preferably, the fault knowledge base includes fault phenomena, fault causes, fault locations, fault levels, handling measures, and preventive measures.
[0010] Preferably, in step S551, the collected raw data is cleaned, including missing value filling, outlier removal, and data smoothing.
[0011] Preferably, the time-domain features include mean, variance, peak value, kurtosis, and skewness.
[0012] Preferably, step S56 includes: S561: Establish a multi-dimensional performance indicator system, including control performance indicators, energy consumption indicators, equipment status indicators, and process indicators; Control performance indicators include control accuracy, response time, overshoot, settling time, and steady-state error; energy consumption indicators include energy consumption per unit product, steam consumption, electricity consumption, and water consumption. Equipment status indicators include equipment operating efficiency, equipment load rate, and equipment health. Process indicators include product yield, product quality, and wastewater treatment effectiveness; Set reasonable weight coefficients for each indicator. The weight coefficients are determined based on the importance and economy of the process, and the weights are calculated using the analytic hierarchy process or the entropy weight method. S562: Real-time acquisition of process parameters such as steam compressor speed, forced circulation pump flow rate, feed rate, temperature, pressure, liquid level, flow rate, power consumption, steam flow rate, product output, and wastewater treatment volume; data cleaning of the acquired raw data, including outlier removal, data smoothing, and missing value imputation. S563: A performance evaluation model is established using a multi-index comprehensive evaluation method; S564: Establish a mapping model between control parameters and performance indicators; S565: Establish short-term, medium-term, and long-term forecasting models; S566: By adopting an adaptive control model, the controller parameters are automatically adjusted when system parameters change.
[0013] Preferably, in step S565, the short-term prediction model predicts the trend of process parameter changes over the next 30 to 60 minutes using time series analysis methods, including ARIMA model, exponential smoothing, and long short-term memory network, to establish a medium-term prediction model to predict load changes over the next 1 to 8 hours. A long-term prediction model is established using support vector regression, random forest regression, and gradient boosting tree to predict load changes over the next 1 to 7 days. Combining historical data, weather forecasts, and production plans, a model predictive control method is used to solve a finite-time domain optimization problem in each control cycle to obtain the optimal control sequence. Based on peak and off-peak electricity prices, an optimal operating plan is formulated, increasing the load during off-peak hours and decreasing the load during peak hours.
[0014] This invention provides a zero-discharge treatment method for methylnitroguanidine wastewater using ultra-low temperature MVR coupled with reverse Carnot cycle, which has the following beneficial effects: 1. The ultra-low temperature MVR coupled reverse Carnot cycle zero-discharge treatment method for methyl nitroguanidine wastewater completely separates organic matter and salts in methyl nitroguanidine wastewater through the coupling process of ultra-low temperature MVR evaporation and concentration with reverse Carnot cycle DTB crystallization, achieving the goal of complete resource utilization and zero discharge of wastewater. The entire treatment process generates no secondary pollution and the treatment effect is stable and reliable.
[0015] 2. The ultra-low temperature MVR coupled reverse Carnot cycle zero-discharge treatment method for methyl nitroguanidine wastewater constructs a load prediction model, an energy consumption optimization model, and a fault diagnosis model. It adopts multi-timescale predictive control to achieve wide-load adaptive adjustment. It can automatically adjust the steam compressor speed, forced circulation pump flow rate, and feed rate according to real-time influent load, water quality changes, and equipment status. Combined with peak and off-peak electricity prices, it formulates the optimal operation plan to ensure stable system operation. Attached Figure Description
[0016] Figure 1 This is a schematic diagram of the method structure of the present invention. Detailed Implementation
[0017] 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.
[0018] Please see Figure 1 This invention provides a technical solution: a method for zero-discharge treatment of methylnitroguanidine wastewater using ultra-low temperature MVR coupled with reverse Carnot cycle, comprising the following steps: Step 1: After the methyl nitroguanidine wastewater enters the equalization tank for homogenization and equalization, it passes through an oil removal device to remove floating oil, a multi-media filter to remove suspended solids, a stripping device to remove volatile organic compounds, and an ozone catalytic oxidation-aerated biological filter to degrade organic matter. Step 2: The pretreated wastewater is fed into the MVR evaporator for evaporation and concentration under ultra-low temperature conditions. The MVR evaporator compresses and heats the secondary steam and then uses it as a heat source for recycling. The concentrated liquid is then fed into a heated separator for further separation. Step 3: The concentrate from Step 2 is fed into the DTB crystallizer, where it undergoes low-temperature evaporation and crystallization under the action of the reverse Carnot cycle system. The refrigerant is compressed and heated by the compressor, releasing heat in the refrigerant condenser. After being depressurized by the expansion valve, it absorbs heat in the refrigerant evaporator. Step 4: After centrifugation, the crystalline product obtained in Step 3 is dried in a drying oven to obtain a solid salt product; Step 5: The intelligent adjustment system automatically adjusts the steam compressor speed, forced circulation pump flow rate, and feed rate according to load changes.
[0019] Step five, which involves using an intelligent adjustment system to automatically adjust the steam compressor speed, forced circulation pump flow rate, and feed rate based on load changes, includes: S51: Establish a real-time data acquisition system to collect process parameters, equipment parameters, environmental parameters, and energy consumption parameters; perform data cleaning and feature extraction; S52: Construct load forecasting models, energy consumption optimization models, and fault diagnosis models; S53: Deploy the model on an edge computing gateway and use incremental learning algorithms to continuously optimize the model; S54: Employs multi-timescale predictive control to achieve wide-load adaptive adjustment and formulates optimal operation plans in conjunction with peak-valley electricity pricing; S55: Employ a combination of unsupervised and supervised learning methods for anomaly detection, predict the remaining lifespan of critical equipment, and build an equipment fault knowledge base. S56: Real-time monitoring of performance indicators, automatic adjustment of control parameters based on evaluation results, establishment of a data closed-loop mechanism for continuous optimization, real-time monitoring of control accuracy, response time, stability and energy consumption indicators, setting of evaluation thresholds, obtaining performance evaluation scores through performance evaluation, and finding the optimal combination of control parameters when the performance evaluation score is lower than the threshold, automatically adjusting the steam compressor speed, forced circulation pump flow rate and feed rate control parameters.
[0020] S55: Employs a combination of unsupervised and supervised learning methods for anomaly detection, predicts the remaining lifespan of critical equipment, and constructs an equipment fault knowledge base, specifically including: S551: Collect equipment operation data, process parameters, environmental parameters and energy consumption data, preprocess the data using standardization or normalization methods, store the preprocessed data in a time series database, and establish a data tagging system, including normal state, abnormal state and fault type tags. S552: Extract time-domain features, frequency-domain features, and time-frequency-domain features from the original data. Use principal component analysis and linear discriminant analysis to perform feature dimensionality reduction, retain principal components with the required contribution rate, and use recursive feature elimination and tree-based feature importance evaluation methods to select features and screen out a subset of features that are highly correlated with the equipment status. S553: Density-based clustering algorithm is used to perform cluster analysis on normal operating condition data to establish normal operating condition data clusters. Anomaly samples are detected by the isolated forest algorithm, and the anomaly score of the samples is calculated. An autoencoder is used for unsupervised feature learning. Anomalies are detected by reconstruction error. The local anomaly factor algorithm is used to calculate the local density of the samples and identify local anomaly points. The detection results of multiple unsupervised detection methods are fused to obtain the final anomaly detection result. S554: Construct a fault diagnosis sample library, including normal samples and various fault samples. Fault types include mechanical faults, electrical faults, thermal faults, and control faults. Use machine learning algorithms such as support vector machines, random forests, and gradient boosting trees to establish a multi-class fault diagnosis model. S555: Establish an equipment degradation model, using a physical model-based approach, combining equipment operating parameters and physical laws to establish degradation equations, and using a data-driven approach to extract equipment degradation features, including trend features, fluctuation features, and abrupt change features. Using machine learning-based methods, including support vector regression, random forest regression, and neural network regression, establish a remaining life prediction model, and using deep learning-based methods, including long short-term memory networks and gated recurrent units, establish a sequence prediction model to predict the remaining life of the equipment.
[0021] The fault knowledge base includes fault phenomena, fault causes, fault locations, fault levels, handling measures, and preventive measures.
[0022] In step S551, the collected raw data is cleaned, including missing value imputation, outlier removal, and data smoothing.
[0023] Among them, the time-domain features include mean, variance, peak value, kurtosis, and skewness.
[0024] Step S56 includes: S561: Establish a multi-dimensional performance indicator system, including control performance indicators, energy consumption indicators, equipment status indicators, and process indicators; Control performance indicators include control accuracy, response time, overshoot, settling time, and steady-state error; energy consumption indicators include energy consumption per unit product, steam consumption, electricity consumption, and water consumption. Equipment status indicators include equipment operating efficiency, equipment load rate, and equipment health. Process indicators include product yield, product quality, and wastewater treatment effectiveness; Set reasonable weight coefficients for each indicator. The weight coefficients are determined based on the importance and economy of the process, and the weights are calculated using the analytic hierarchy process or the entropy weight method. S562: Real-time acquisition of process parameters such as steam compressor speed, forced circulation pump flow rate, feed rate, temperature, pressure, liquid level, flow rate, power consumption, steam flow rate, product output, and wastewater treatment volume; data cleaning of the acquired raw data, including outlier removal, data smoothing, and missing value imputation. S563: A performance evaluation model is established using a multi-index comprehensive evaluation method; S564: Establish a mapping model between control parameters and performance indicators; S565: Establish short-term, medium-term, and long-term forecasting models; S566: By adopting an adaptive control model, the controller parameters are automatically adjusted when the system parameters change; In step S565, a short-term forecasting model predicts the trend of process parameter changes over the next 30 to 60 minutes using time series analysis methods, including ARIMA model, exponential smoothing, and long short-term memory networks. A medium-term forecasting model is then established to predict load changes over the next 1 to 8 hours using support vector regression, random forest regression, and gradient boosting trees. Finally, a long-term forecasting model is established to predict load changes over the next 1 to 7 days. Combining historical data, weather forecasts, and production plans, a model predictive control method is used to solve a finite-time domain optimization problem in each control cycle to obtain the optimal control sequence. Based on peak and off-peak electricity prices, an optimal operating plan is formulated to increase the load during off-peak hours and decrease the load during peak hours.
[0025] In summary, this method for zero-discharge treatment of methylnitroguanidine wastewater using ultra-low temperature MVR coupled with reverse Carnot cycle is effective when in use. Example 1: Treatment of Methylnitroguanidine Wastewater The treatment capacity is 60 m³ / h for methylnitroguanidine production wastewater with a COD concentration of 10,000–15,000 mg / L, a total nitrogen concentration of 1,800–2,500 mg / L, and a salt content of 4–6%.
[0026] The equipment includes: equalization tank, oil removal device, multi-media filter, stripping tower, ozone catalytic oxidation device, aerated biological filter, MVR evaporator, DTB crystallizer, centrifuge, and drying oven; S1: Methylnitroguanidine wastewater is discharged into an equalization tank for homogenization and flow equalization. The equalized wastewater then passes through an oil removal device to remove floating oil, followed by a multi-media filter to remove suspended solids, and then a stripping device to remove volatile organic compounds (VOCs). The stripping temperature is controlled at 45–55℃, the air-to-water ratio is 18:1, and the VOC removal rate is ≥88%. Subsequently, the wastewater undergoes a combined ozone catalytic oxidation-aerated biological filter process to degrade organic matter. The ozone dosage is 100 mg / L, the catalytic oxidation reaction time is 2.5 hours, the hydraulic retention time in the aerated biological filter is 5 hours, and the COD removal rate is ≥92%, and the total nitrogen removal rate is ≥85%. S2: The pretreated wastewater enters the MVR evaporator for evaporation and concentration under ultra-low temperature conditions. The MVR evaporator compresses the secondary steam to 75–85°C using a steam compressor and then recycles it as a heat source, with a steam compression ratio of 1.8–2.2. The concentration ratio is 12–18 times. The concentrated liquid enters a heated separator for further separation, with the separation temperature controlled at 65–75°C. S3: The concentrated liquid is discharged into the DTB crystallizer, where low-temperature evaporation crystallization is carried out under the action of the reverse Carnot cycle system. R410a refrigerant is used. After being compressed and heated to 65–75°C by the compressor, heat is released in the refrigerant condenser. After being depressurized to -12 to -8°C by the expansion valve, heat is absorbed in the refrigerant evaporator. The crystallization temperature is controlled at -8 to -2°C, the crystallization time is 10–14 hours, and the crystallization yield is ≥96%. S4: The crystallized product is separated by centrifugation at a speed of 1800–2200 rpm for 25 minutes to obtain a wet salt product. The wet salt is then dried in a drying oven at a temperature of 85–105℃ for 1.5–3 hours to obtain a solid salt product with a moisture content of ≤0.3% and a methylnitroguanidine content of ≤0.05%, which meets the industrial salt standard.
[0027] Example 2: In Example 1, an intelligent adjustment system is used to automatically adjust the steam compressor speed, forced circulation pump flow rate, and feed rate according to load changes; S1: Methylnitroguanidine wastewater is discharged into the equalization tank. The intelligent adjustment system predicts the load changes of subsequent units based on the liquid level of the equalization tank and the online monitoring value of COD of the influent, and sends a feedforward signal to the pretreatment section. The subsequent stripping device will automatically fine-tune the steam flow and air inlet valve according to the concentration of volatile organic compounds in the influent to dynamically stabilize the gas-water ratio in the optimal range of 18:1. The ozone generator of the ozone catalytic oxidation device will automatically adjust the ozone output according to the wastewater flow rate and real-time COD value entering the unit to ensure oxidation efficiency and avoid reagent waste. S2: The pretreated wastewater enters the MVR evaporator. Based on the liquid level, density, and boiling point elevation data within the evaporator, the frequency of the feed pump is automatically adjusted to maintain material balance and prevent the concentrate from becoming too thick or the evaporation rate from being insufficient. Compressor speed regulation: The steam compressor speed is no longer fixed but dynamically adjusted based on the temperature and pressure of the secondary steam and the set concentration ratio target (12-18 times). When the feed concentration decreases, the speed is automatically reduced to save energy; when it is necessary to increase the evaporation intensity, the speed is increased to ensure the heat source temperature is maintained within the effective range of 75-85℃. The forced circulation pump flow rate is linked to the compressor speed and the crystallization trend within the evaporator. When a risk of scaling is detected, the circulation flow rate is automatically increased to enhance the flushing effect; during stable operation, the flow rate is reduced to the economic flow rate. S3: The concentrate enters the DTB crystallizer. The intelligent adjustment system monitors key parameters such as temperature, suspension density, and supersaturation (calculated indirectly by online particle size analyzer or temperature difference method) in the crystallizer to achieve precise control. The system adjusts the power of the refrigerant compressor and the opening of the expansion valve to precisely control the crystallization temperature to be stable within the set range of -8 to -2℃. Based on the crystal growth rate and product particle size distribution, the system automatically optimizes the flow rate of the circulation pump and the opening cycle of the fine crystal elimination unit to ensure that the crystallization yield is stable at a high level of ≥96%. S4: Linked control of product separation and drying. The crystallized slurry enters the centrifuge and drying chamber. The centrifuge speed (1800–2200 rpm) and separation time can be finely adjusted according to the concentration and particle size of the feed slurry to ensure the best separation effect and low wet salt moisture content. The drying temperature (85–105℃) and time of the drying chamber are set. The online humidity detection signal from the centrifuge wet salt is received to achieve dynamic adjustment, ultimately ensuring that the solid salt product moisture content is ≤0.3% and the methylnitroguanidine content is ≤0.05%.
[0028] All standard parts used in this application can be purchased from the market, and can be customized according to the description and drawings. The specific connection methods of each part adopt conventional methods such as bolts, rivets, and welding that are mature in the prior art. The machinery, parts and equipment all adopt conventional models in the prior art. The installation methods between equipment are also the same as conventional installation methods in the prior art. For example, the two ends of shaft-shaped parts are connected by bearings, the connection position of valve components is provided with anti-leakage rubber strips, the outside of threaded rods or lead rods is provided with dust covers, and the equipment can be driven by either built-in batteries or external power supply. The control method is automatic control by a controller. The control circuit of the controller can be implemented by simple programming by those skilled in the art and is common knowledge in the field. Since this invention is mainly used to protect mechanical devices, this invention will not explain the control method and circuit connection in detail. The external controller mentioned in the specification can play a control role for the electrical components mentioned herein, and the external controller is a conventional known device.
[0029] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.
Claims
1. A method for zero-discharge treatment of methylnitroguanidine wastewater using ultra-low temperature MVR coupled with reverse Carnot cycle, characterized in that: Includes the following steps: Step 1: After the methyl nitroguanidine wastewater enters the equalization tank for homogenization and equalization, it passes through an oil removal device to remove floating oil, a multi-media filter to remove suspended solids, a stripping device to remove volatile organic compounds, and an ozone catalytic oxidation-aerated biological filter to degrade organic matter. Step 2: The pretreated wastewater is fed into the MVR evaporator for evaporation and concentration under ultra-low temperature conditions. The MVR evaporator compresses and heats the secondary steam and then uses it as a heat source for recycling. The concentrated liquid is then fed into a heated separator for further separation. Step 3: The concentrate from Step 2 is fed into the DTB crystallizer, where it undergoes low-temperature evaporation and crystallization under the action of the reverse Carnot cycle system. The refrigerant is compressed and heated by the compressor, releasing heat in the refrigerant condenser. After being depressurized by the expansion valve, it absorbs heat in the refrigerant evaporator. Step 4: After centrifugation, the crystalline product obtained in Step 3 is dried in a drying oven to obtain a solid salt product; Step 5: The intelligent adjustment system automatically adjusts the steam compressor speed, forced circulation pump flow rate, and feed rate according to load changes.
2. The method for zero-discharge treatment of methylnitroguanidine wastewater using ultra-low temperature MVR coupled with reverse Carnot cycle according to claim 1, characterized in that: Step five, the method of automatically adjusting the steam compressor speed, forced circulation pump flow rate, and feed rate according to load changes through an intelligent adjustment system, includes: S51: Establish a real-time data acquisition system to collect process parameters, equipment parameters, environmental parameters, and energy consumption parameters; perform data cleaning and feature extraction; S52: Construct load forecasting models, energy consumption optimization models, and fault diagnosis models; S53: Deploy the model on an edge computing gateway and use incremental learning algorithms to continuously optimize the model; S54: Employs multi-timescale predictive control to achieve wide-load adaptive adjustment and formulates optimal operation plans in conjunction with peak-valley electricity pricing; S55: Employ a combination of unsupervised and supervised learning methods for anomaly detection, predict the remaining lifespan of critical equipment, and build an equipment fault knowledge base. S56: Real-time monitoring of performance indicators, automatic adjustment of control parameters based on evaluation results, establishment of a data closed-loop mechanism for continuous optimization, real-time monitoring of control accuracy, response time, stability and energy consumption indicators, setting of evaluation thresholds, obtaining performance evaluation scores through performance evaluation, and finding the optimal combination of control parameters when the performance evaluation score is lower than the threshold, automatically adjusting the steam compressor speed, forced circulation pump flow rate and feed rate control parameters.
3. The method for zero-discharge treatment of methylnitroguanidine wastewater using ultra-low temperature MVR coupled with reverse Carnot cycle according to claim 2, characterized in that: S55: Employing a combination of unsupervised and supervised learning methods for anomaly detection, predicting the remaining lifespan of critical equipment, and constructing an equipment fault knowledge base, specifically including: S551: Collect equipment operation data, process parameters, environmental parameters and energy consumption data, preprocess the data using standardization or normalization methods, store the preprocessed data in a time series database, and establish a data tagging system, including normal state, abnormal state and fault type tags. S552: Extract time-domain features, frequency-domain features, and time-frequency-domain features from the raw data; S553: Density-based clustering algorithm is used to perform cluster analysis on normal operating condition data to establish normal operating condition data clusters. Anomaly samples are detected by the isolated forest algorithm, and the anomaly score of the samples is calculated. An autoencoder is used for unsupervised feature learning. Anomalies are detected by reconstruction error. The local anomaly factor algorithm is used to calculate the local density of the samples and identify local anomaly points. The detection results of multiple unsupervised detection methods are fused to obtain the final anomaly detection result. S554: Construct a fault diagnosis sample library, including normal samples and various fault samples. Fault types include mechanical faults, electrical faults, thermal faults, and control faults. Use machine learning algorithms such as support vector machines, random forests, and gradient boosting trees to establish a multi-class fault diagnosis model. S555: Establish an equipment degradation model, using a physical model-based approach, combining equipment operating parameters and physical laws to establish degradation equations, using a data-driven approach to extract equipment degradation features, using machine learning-based methods, including support vector regression, random forest regression, and neural network regression, to establish a remaining life prediction model, and using deep learning-based methods, including long short-term memory networks and gated recurrent units, to establish a sequence prediction model to predict the remaining life of the equipment.
4. The method for zero-discharge treatment of methylnitroguanidine wastewater using ultra-low temperature MVR coupled with reverse Carnot cycle according to claim 1, characterized in that: The fault knowledge base includes fault phenomena, fault causes, fault locations, fault levels, handling measures, and preventive measures.
5. The method for zero-discharge treatment of methylnitroguanidine wastewater using ultra-low temperature MVR coupled with reverse Carnot cycle according to claim 3, characterized in that: In step S551, the collected raw data is cleaned, including missing value filling, outlier removal, and data smoothing.
6. The method for zero-discharge treatment of methylnitroguanidine wastewater using ultra-low temperature MVR coupled with reverse Carnot cycle according to claim 3, characterized in that: The time-domain features include mean, variance, peak value, kurtosis, and skewness.
7. The method for zero-discharge treatment of methylnitroguanidine wastewater using ultra-low temperature MVR coupled with reverse Carnot cycle according to claim 2, characterized in that: Step S56 includes: S561: Establish a multi-dimensional performance indicator system, including control performance indicators, energy consumption indicators, equipment status indicators, and process indicators; Control performance indicators include control accuracy, response time, overshoot, settling time, and steady-state error; energy consumption indicators include energy consumption per unit product, steam consumption, electricity consumption, and water consumption. Equipment status indicators include equipment operating efficiency, equipment load rate, and equipment health. Process indicators include product yield, product quality, and wastewater treatment effectiveness; Set reasonable weight coefficients for each indicator. The weight coefficients are determined based on the importance and economy of the process, and the weights are calculated using the analytic hierarchy process or the entropy weight method. S562: Real-time acquisition of process parameters such as steam compressor speed, forced circulation pump flow rate, feed rate, temperature, pressure, liquid level, flow rate, power consumption, steam flow rate, product output, and wastewater treatment volume; data cleaning of the acquired raw data, including outlier removal, data smoothing, and missing value imputation. S563: A performance evaluation model is established using a multi-index comprehensive evaluation method; S564: Establish a mapping model between control parameters and performance indicators; S565: Establish short-term, medium-term, and long-term forecasting models; S566: By adopting an adaptive control model, the controller parameters are automatically adjusted when system parameters change.
8. The method for zero-discharge treatment of methylnitroguanidine wastewater using ultra-low temperature MVR coupled with reverse Carnot cycle according to claim 7, characterized in that: In step S565, the short-term prediction model predicts the trend of process parameter changes over the next 30 to 60 minutes. It uses time series analysis methods, including ARIMA model, exponential smoothing, and long short-term memory network, to establish a medium-term prediction model to predict load changes over the next 1 to 8 hours. It uses support vector regression, random forest regression, and gradient boosting tree to establish a long-term prediction model to predict load changes over the next 1 to 7 days. Combining historical data, weather forecasts, and production plans, it uses model predictive control methods to solve a finite-time domain optimization problem in each control cycle to obtain the optimal control sequence. Combined with peak and off-peak electricity prices, it formulates an optimal operation plan to increase the load during off-peak hours and decrease the load during peak hours.