Ozone treatment device and method of operation thereof

Abstract

The application discloses an ozone treatment device and an operation method thereof, and belongs to the technical field of wastewater advanced oxidation treatment. The ozone treatment device comprises a customized ozone oxidation tower, an ozone catalyst, an ozone generating device, an ozone concentration detection unit, an online water inlet pollutant concentration detection unit and an AI intelligent control unit. The oxidation tower optimizes the height-diameter ratio and the water distribution and air distribution structure and is internally provided with the ozone catalyst. The AI intelligent control unit receives water inlet quality data in real time, dynamically calculates the optimal ozone dosage by using a pre-trained prediction model, and realizes precise closed-loop dosing by real-time regulation and control of the ozone generating device. The catalytic oxidation tower is combined to strengthen mass transfer and catalytic reaction, and the model is self-optimized through water outlet data feedback. The application significantly improves the ozone utilization rate and oxidation efficiency, and reduces the system operation energy consumption and treatment cost.

Description

Technical Field

[0001] This invention relates to the field of advanced wastewater oxidation treatment technology, specifically to an ozone treatment device and its operation method. Background Technology

[0002] Ozone oxidation technology, as a typical advanced oxidation process, has been widely used in the deep treatment of recalcitrant organic wastewater in industries such as chemical and dyeing due to its strong oxidation capacity and lack of secondary pollution residues. However, in practical engineering applications, existing ozone oxidation processes still have many limitations: traditional ozone oxidation reactors mostly adopt a general bubble column structure, resulting in limited gas-liquid two-phase contact efficiency, slow mass transfer rate of ozone in water, and a large amount of unreacted ozone escaping with the exhaust gas, which not only wastes ozone resources but also increases the load and cost of exhaust gas treatment. At the operation and control level, current ozone dosing generally adopts an experience-based setting method based on the influent water quality range, or a control mode with fixed ozone concentration and fixed dosing ratio. This cannot dynamically respond to real-time fluctuations in influent COD, UV254, and characteristic pollutant concentrations, easily leading to problems such as incomplete oxidation and difficulty in consistently meeting effluent water quality standards when influent pollutant concentrations are high, and excessive ozone dosing when influent pollutant concentrations are low, resulting in high system operating costs. Meanwhile, existing catalytic ozone technologies often directly fill the catalyst into a general-purpose reactor without considering the characteristics of the catalytic reaction and matching the reactor's height-to-diameter ratio, gas distribution structure, and catalyst loading method. As a result, the catalytic efficiency of the catalyst cannot be fully utilized, and most control systems lack an effluent feedback self-correction mechanism. They cannot automatically optimize control parameters as water quality changes and operating conditions deteriorate, making it difficult to maintain optimal operating conditions in the long term. This restricts the improvement of ozone oxidation process efficiency and its large-scale application. Summary of the Invention

[0003] To address the shortcomings of existing technologies, the present invention aims to provide an ozone treatment device and its operation method, which establishes a dynamic nonlinear mapping relationship between the influent pollutant load and the optimal ozone dosage through an AI intelligent control unit, thereby achieving precise dosing on demand.

[0004] To achieve the above objectives, the embodiments of this invention provide the following technical solutions:

[0005] This application provides an operating method for an ozone treatment device, including: S1, system initialization: loading a pre-trained ozone dosage prediction model in the AI ​​intelligent control unit; S2, data acquisition: real-time acquisition of influent water quality data through an online influent pollutant concentration detection unit, and transmitting the influent water quality data to the AI ​​intelligent control unit; S3, intelligent prediction and control: the AI ​​intelligent control unit calculates the instantaneous optimal ozone dosage based on the real-time influent water quality data, current process parameters, and target effluent requirements using the ozone dosage prediction model, and generates a control signal; S4, ozone dosing: the control signal drives the ozone generator to introduce precisely measured ozone into the customized... Bottom of the ozone oxidation tower; S5, Catalytic oxidation reaction: Ozone is distributed into microbubbles through the gas distribution structure in the customized ozone oxidation tower, flowing upward from the bottom of the tower and contacting the wastewater distributed downward from the top of the tower in a counter-current manner. Under the action of the ozone catalyst in the tower, a catalytic oxidation reaction occurs. The treated water is discharged from the outlet at the bottom of the tower, and the tail gas is discharged into the atmosphere after being treated by the tail gas destruction and collection device at the top; S6, Feedback self-correction: The effluent water quality data is periodically collected by the online effluent water quality detection unit and fed back to the AI ​​intelligent control unit. The effluent water quality detection results are compared with the model prediction results. When the deviation continues to exceed the set threshold, the parameters of the ozone dosage prediction model are automatically optimized and updated.

[0006] Furthermore, the height-to-diameter ratio of the customized ozone oxidation tower is 5~15:1; a microporous aeration disc is provided at the bottom of the tower, and an ozone catalyst packing layer is filled in the middle, with guide plates provided between each layer of the ozone catalyst packing layer; a Pall ring packing layer is provided at the top of the tower, and the Pall ring packing layer is located between the ozone catalyst packing layer and the water distributor at the top of the tower; a tail gas destruction and collection device is provided at the top of the tower; and multiple sampling ports and observation windows are provided on the side walls of the tower.

[0007] Further, in step S2, the online detection unit for pollutant concentration in the influent is installed on the influent pipe of the system to detect the concentration of organic matter in the influent in real time. The organic matter concentration indicators include at least one of COD value, UV254 value, and TOC value. In step S4, an ozone concentration detection unit is installed on the air inlet pipe of the customized ozone oxidation tower to monitor the ozone dosage concentration.

[0008] Furthermore, the ozone dosage prediction model built into the AI ​​intelligent control unit is a machine learning model, which includes gradient boosting decision tree, random forest or neural network model; the input features of the machine learning model include at least real-time influent COD value or real-time UV254 value, influent flow rate, pH value and water temperature, and the output is the optimal ozone dosage flow rate or optimal ozone dosage concentration.

[0009] Furthermore, the ozone catalyst in the ozone catalyst packing layer is a catalytic packing material supported on one of Raschig ring packing, honeycomb ceramic, porous alumina or modified activated carbon, and loaded with at least one active component selected from Mn, Cu, Fe, Co and Ni; the loading of the active component of the ozone catalyst is 2% to 10%.

[0010] Furthermore, in S6, the AI ​​intelligent control unit periodically collects effluent water quality data through the online effluent water quality detection unit, compares the effluent pollutant concentration detection results with the effluent pollutant concentration predicted by the ozone dosage prediction model, and automatically triggers the model parameter fine-tuning program when the deviation continues to exceed the set threshold to optimize and update the parameters of the ozone dosage prediction model, thereby realizing online feedback learning of the prediction model.

[0011] Accordingly, this application also provides an ozone treatment device, comprising: a customized ozone oxidation tower with a height-to-diameter ratio of 5 to 15:1, a microporous aeration disc at the bottom of the tower, an ozone catalyst packing layer in the middle with guide plates between the layers, a Pall ring packing layer at the top, a tail gas destruction and collection device at the top, and multi-stage sampling ports and observation windows on the side walls; an ozone generator connected to the air inlet at the bottom of the customized ozone oxidation tower for supplying ozone into the tower; an ozone concentration detection unit installed on the air inlet pipe between the ozone generator and the customized ozone oxidation tower for monitoring the ozone dosage concentration; and an online detection unit for influent pollutant concentration installed on the system influent pipe for monitoring the ozone concentration. The system includes a real-time influent water quality monitoring unit; an effluent water quality online monitoring unit, installed on the system's effluent pipeline, for real-time influent water quality data collection; an AI intelligent control unit, whose signal input terminals are connected to the influent pollutant concentration online detection unit and the ozone concentration detection unit, respectively, and whose control output terminal is connected to the ozone generator; the AI ​​intelligent control unit has a built-in pre-trained ozone dosage prediction model, used to calculate the instantaneous optimal ozone dosage based on real-time influent water quality data, and output control signals to the ozone generator; the AI ​​intelligent control unit also has a feedback learning module, used to automatically optimize and update the parameters of the prediction model based on the effluent water quality detection results.

[0012] Furthermore, the customized ozone oxidation tower also includes a circulation pump and a booster pump; the inlet and outlet of the circulation pump are connected to the lower and upper parts of the tower body respectively through circulation pipes, so as to circulate and mix the sewage in the tower; the booster pump is connected to the outlet at the bottom of the tower body, so as to discharge the treated water from the bottom of the tower.

[0013] Furthermore, the ozone catalyst in the ozone catalyst packing layer is a honeycomb ceramic catalyst supported on MnO2-CeO2; the microporous aeration disc is a titanium alloy microporous aeration disc.

[0014] The beneficial effects of this invention are as follows: A dynamic nonlinear mapping relationship between the influent pollutant load and the optimal ozone dosage is established through an AI intelligent control unit, enabling precise dosing on demand; the integrated structural design of the customized ozone oxidation tower enhances the gas-liquid-solid three-phase mass transfer; and a highly efficient heterogeneous ozone catalyst promotes ozone decomposition to generate a large number of hydroxyl radicals. The synergistic effect of these three factors significantly improves ozone utilization and pollutant removal efficiency, and greatly reduces operating costs. Attached Figure Description

[0015] Figure 1 A schematic flowchart illustrating the operation method of an ozone treatment device provided in this application embodiment;

[0016] Figure 2 This is a schematic diagram of an ozone treatment device provided in an embodiment of this application.

[0017] Attached reference numerals: 1-Exhaust gas destruction and collection device, 2-Water distributor, 3-Pall ring packing layer, 4-Ozone catalyst packing layer, 5-Microporous aeration disc, 6-Ozone inlet pipe, 7-Circulation pipe, 8-Water inlet pipe, 9-Water outlet pipe. Detailed Implementation

[0018] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0019] In this invention, the terms "system" and "network" are used interchangeably. "Multiple" refers to two or more; therefore, in this invention, "multiple" can also be understood as "at least two." "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Additionally, the character " / ", unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship. Furthermore, it should be understood that in the description of this invention, terms such as "first" and "second" are used only for descriptive purposes and should not be construed as indicating or implying relative importance or order.

[0020] Example 1:

[0021] like Figure 1As shown in the figure, this application provides an operation method for an ozone treatment device, including: S1, system initialization: loading a pre-trained ozone dosage prediction model in the AI ​​intelligent control unit; S2, data acquisition: real-time acquisition of influent water quality data through an online influent pollutant concentration detection unit, and transmitting the influent water quality data to the AI ​​intelligent control unit; S3, intelligent prediction and control: the AI ​​intelligent control unit calculates the instantaneous optimal ozone dosage based on the real-time influent water quality data, current process parameters, and target effluent requirements through the ozone dosage prediction model, and generates a control signal; S4, ozone dosing: the control signal drives the ozone generator to introduce precisely measured ozone into the system. S5. Catalytic oxidation reaction: Ozone is distributed into microbubbles in the customized ozone oxidation tower through the gas distribution structure. It flows upward from the bottom of the tower and comes into counter-current contact with the sewage distributed downward from the top of the tower. Under the action of the ozone catalyst in the tower, a catalytic oxidation reaction occurs. The treated water is discharged from the outlet at the bottom of the tower. The tail gas is treated by the tail gas destruction and collection device 1 at the top and then discharged into the atmosphere. S6. Feedback self-correction: The effluent water quality data is periodically collected by the online effluent water quality detection unit and fed back to the AI ​​intelligent control unit. The effluent water quality detection results are compared with the model prediction results. When the deviation continues to exceed the set threshold, the parameters of the ozone dosage prediction model are automatically optimized and updated.

[0022] In one possible embodiment, the object of treatment in this embodiment is the biochemical effluent from a comprehensive wastewater treatment plant in a chemical industrial park, with a designed treatment capacity of 10 m³ / h. First, system initialization is performed. A pre-trained ozone dosage prediction model is loaded into the industrial computer-type AI intelligent control unit, and communication verification between each detection unit, ozone generator, and control unit is completed to ensure stable signal transmission. Then, the influent system is started, with wastewater entering the system at a flow rate of 10 m³ / h. Influent water quality data is collected in real time by an online pollutant concentration detection unit installed on the influent pipe 8, and the instantaneous UV254 value is measured to be 0.35 cm⁻¹. -1Simultaneously, parameters such as influent flow rate and pH value are collected, and all data are transmitted in real time to the AI ​​intelligent control unit via signal lines. The AI ​​intelligent control unit then receives the real-time influent water quality data, combines it with the current system process parameters and the preset target requirement of effluent COD ≤ 40 mg / L, and calls the built-in ozone dosage prediction model to calculate the instantaneous optimal ozone dosage under the current operating conditions as 70 mg / L, corresponding to an ozone dosage of 700 g per hour, and generates a corresponding ozone production control signal. Subsequently, the control signal is transmitted to the ozone generator, driving the ozone generator to adjust its operating power and airflow to precisely control the ozone production at 700 g / h, and the precisely metered ozone is introduced into the bottom of the customized ozone oxidation tower through the air inlet valve. After entering the tower, the ozone is dispersed into microbubbles by the gas distribution structure, flows upward from the bottom of the tower, and forms a counter-current flow with the sewage distributed downward from the top of the tower to fully contact it, and the ozone catalyst filled in the tower is also present. The reaction induces the generation of a large number of hydroxyl radicals (・OH), which deeply oxidize the organic pollutants in the wastewater. After the reaction is completed, the treated water is discharged from the bottom of the tower. The COD of the effluent can be reduced to below 40 mg / L. The remaining ozone tail gas rises to the top of the tower and is discharged after being decomposed and treated by the tail gas destruction and collection device 1 to meet the standards. Finally, during the operation of the system, the COD detection data of the effluent from the ozone oxidation tower is collected by the online effluent water quality detection unit set on the effluent pipe 9 every 4 hours. The data is fed back to the AI ​​intelligent control unit. The actual effluent COD detection result is compared with the preset target water quality requirement of effluent COD ≤ 40 mg / L. When the deviation between the two exceeds the preset threshold, the system automatically starts the model parameter fine-tuning program. The incremental learning method is used to update the model weights based on the latest operating data and optimize and correct the parameters of the ozone dosage prediction model, so as to realize the online self-learning and iterative optimization of the model.

[0023] By combining AI prediction models with real-time influent water quality dynamic calculation of the optimal ozone dosage, and with the support of a closed-loop control system for the entire process of catalytic oxidation reaction and effluent feedback self-correction, the ozone dosage is dynamically adjusted on demand and responded in seconds. This breaks through the extensive mode of traditional fixed ratio or experience-based dosage, ensuring stable effluent water quality while avoiding excessive ozone dosage, thus improving ozone utilization efficiency from the perspective of operational strategy.

[0024] In this embodiment, the height-to-diameter ratio of the customized ozone oxidation tower is 5~15:1; a microporous aeration disc 5 is provided at the bottom of the tower, and an ozone catalyst packing layer 4 is filled in the middle, with guide plates provided between each layer of the ozone catalyst packing layer 4; a Pall ring packing layer 3 is provided at the top of the tower, and the Pall ring packing layer 3 is located between the ozone catalyst packing layer 4 and the water distributor 2 at the top of the tower; a tail gas destruction and collection device 1 is provided at the top of the tower; and multi-stage sampling ports and observation windows are provided on the side wall of the tower.

[0025] In one possible embodiment, a stainless steel ozone oxidation tower body is first custom-made, with a height of 8m, a diameter of 1m, and a height-to-diameter ratio of 8:1, falling within the design range of 5 to 15:1. Then, a titanium alloy microporous aeration disc 5 is installed at the bottom of the tower as an air distribution structure. In the middle region of the tower, ozone catalyst packing is filled in three layers, each 1m high, forming an ozone catalyst packing layer 4. Baffles are installed between each layer of ozone catalyst packing layer 4 to regulate the flow of water and air within the tower, preventing channeling and extending the contact time between the gas, liquid, and solid phases. Finally, Pall ring packing is installed in the upper region of the tower, above the uppermost catalyst packing layer, forming a Pall ring packing layer 3. This Pall ring packing layer 3 is located within the ozone catalyst... Between the ozone catalyst packing layer 4 and the water distributor 2 at the top of the tower, the rising ozone bubbles are further broken up to increase the gas-liquid contact area and enhance the mass transfer effect. Subsequently, an ozone destroyer is installed at the top of the tower as a gas destruction and collection device 1 to decompose and treat the remaining ozone tail gas after the reaction. At the same time, multiple sampling ports and transparent observation windows are set along the height direction on the side wall of the tower to collect water samples at different tower heights for water quality testing and to visually observe the gas-liquid operation status inside the tower. Finally, after the water inlet pipe, water outlet pipe and air inlet pipe of the tower are connected, the sewage is evenly distributed downward from the water distributor 2 at the top of the tower, and the ozone diffuses upward from the aeration plate at the bottom of the tower to form microbubbles, which pass through the ozone catalyst packing layer 4 and the Pall ring packing layer 3 in sequence, and come into counter-current contact with the sewage to complete the catalytic oxidation reaction.

[0026] By limiting the height-to-diameter ratio range of the oxidation tower and equipping it with a multi-stage internal structure consisting of microporous aeration discs 5, layered catalyst packing, guide plates, and Pall ring packing, a customized reactor design for the gas-liquid-solid three-phase reaction of catalytic ozone oxidation was achieved. This effectively enhanced the mass transfer efficiency of ozone, increased the contact probability between ozone and catalysts and pollutants, provided an efficient reaction platform for intelligent and precise dosing, and maximized the efficiency of catalytic oxidation.

[0027] In this embodiment of the application, in step S2, the online detection unit for pollutant concentration in the influent is installed on the influent pipe 8 of the system and is used to detect the concentration of organic matter in the influent in real time. The concentration of organic matter includes at least one of COD value, UV254 value, and TOC value. In step S4, an ozone concentration detection unit is installed on the air inlet pipe of the customized ozone oxidation tower and is used to monitor the ozone dosage concentration.

[0028] In one possible embodiment, firstly, an online pollutant concentration detection unit is installed on the system's inlet water pipe 8. This detection unit integrates an online automatic COD sampling analyzer and an online UV254 detector. The device's detection probe extends into the inlet water pipe 8 to detect the organic matter concentration in the inlet water in real time, specifically including at least one of COD, UV254, and TOC values. The detection data is output in real time in the form of an electrical signal. Then, an ozone concentration detection unit, specifically an ozone concentration analyzer, is installed on the air intake pipe between the ozone generator and the customized ozone oxidation tower. Its detection gas path is connected to the air intake pipe for real-time... The ozone concentration introduced into the oxidation tower is monitored. The signal outputs of both the online pollutant concentration detection unit and the ozone concentration detection unit are connected to the signal input of the AI ​​intelligent control unit to ensure that water quality data and ozone concentration data can be transmitted to the control unit in real time and stably. During system operation, the online pollutant concentration detection unit continuously collects data on influent water quality fluctuations, while the ozone concentration detection unit simultaneously monitors the actual ozone concentration. Both types of data serve as the basis for calculations by the AI ​​control unit. Influent water quality data is used to predict ozone demand, while ozone concentration data is used to verify the deviation between the actual dosage and the commanded value.

[0029] By installing an online pollutant concentration detection unit in the water inlet pipe 8 and an ozone concentration detection unit in the air inlet pipe, real-time monitoring of both water inlet load and ozone dosage is achieved. This provides an accurate input data source for the AI ​​intelligent control unit, ensuring the accuracy of ozone dosage prediction. At the same time, it can verify the deviation between the actual ozone dosage and the command value in real time, improving the reliability of closed-loop control.

[0030] In this embodiment of the application, the ozone dosage prediction model built into the AI ​​intelligent control unit is a machine learning model, which includes gradient boosting decision tree, random forest or neural network model; the input features of the machine learning model include at least real-time influent COD value or real-time UV254 value, influent flow rate, pH value and water temperature, and the output is the optimal ozone dosage flow rate or the optimal ozone dosage concentration.

[0031] In one possible embodiment, a machine learning model is first built in the AI ​​intelligent control unit as an ozone dosage prediction model. The machine learning model can be any of gradient boosting decision trees, random forests, or neural network models. In this embodiment, the gradient boosting regression tree (GBDT) algorithm is used to construct the prediction model. Then, six months of historical operating data of the target wastewater are selected as the training dataset to pre-train the model and determine its input features and output parameters. The model's input features include at least real-time influent COD or real-time UV254 values, influent flow rate, pH value, and water temperature. In this embodiment, the model is selected... The real-time UV254 value, influent flow rate, and pH value are taken as core input features. The model output is the optimal ozone dosing flow rate or the optimal ozone dosing concentration. In this embodiment, the ozone dosing concentration (unit: g / Nm³) is used as the output parameter. The pre-trained model is then loaded into the control program of the AI ​​intelligent control unit to complete the initialization and verification of the model parameters. Finally, when the system is running, the AI ​​intelligent control unit imports the real-time collected input feature data into the prediction model and directly outputs the optimal ozone dosing parameters under the corresponding operating conditions through model calculation, driving the ozone generator to execute the ozone supply of the corresponding output.

[0032] By employing machine learning algorithms to construct an ozone dosage prediction model, a dynamic nonlinear mapping relationship between influent pollutant load and optimal ozone dosage was established. Compared with the traditional linear proportional dosing method, this model can more accurately match the ozone demand under fluctuating influent water quality, further improving the accuracy of dosing control and ozone utilization efficiency.

[0033] In this embodiment, the ozone catalyst in the ozone catalyst packing layer is a catalytic packing material that uses one of Raschig ring packing, honeycomb ceramic, porous alumina or modified activated carbon as a support and loads at least one active component of Mn, Cu, Fe, Co and Ni; the loading of the active component of the ozone catalyst is 2% to 10%.

[0034] In one possible embodiment, an ozone catalyst is first prepared, using Raschig ring packing, honeycomb ceramics, porous alumina, or modified activated carbon as the catalyst support. In this embodiment, honeycomb ceramics are selected as the support because their regular pore structure ensures low pressure drop and uniform water and gas distribution. Then, at least one metal element selected from Mn, Cu, Fe, Co, and Ni is chosen as the active component. In this embodiment, the oxide of Mn and Ce, namely MnO2-CeO2, is selected as the composite active component, utilizing the synergistic effect of the bimetallic elements to enhance catalytic activity. Finally, the active component is loaded onto the honeycomb ceramic support using an equal-volume impregnation method. A manganese-cerium mixed nitrate solution with a total metal ion concentration of 0.5 mol / L was prepared as the impregnation solution, and the impregnation time was 4 h, with the loading of the active component controlled at 2%–10%. After drying at a constant temperature of 105℃ for 12 h, it was placed in a muffle furnace and calcined at 500℃ for 4 h. After natural cooling, the finished ozone catalyst was prepared. Finally, the prepared ozone catalyst was neatly packed into the middle packing area of ​​a customized ozone oxidation tower to form ozone catalyst packing layer 4. When wastewater and ozone flow through the packing layer, a catalytic ozone oxidation reaction occurs at the active sites on the catalyst surface, generating highly oxidizing hydroxyl radicals, thereby achieving deep oxidation of recalcitrant organic matter.

[0035] By selecting ozone catalysts with transition metal active components supported on porous carriers, it is possible to effectively catalyze the decomposition of ozone to generate a large number of hydroxyl radicals, which breaks through the reaction selectivity limitation of ozone oxidation alone and improves the oxidation and mineralization ability of recalcitrant organic matter; at the same time, the loading range of 2% to 10% takes into account both catalytic activity and preparation cost, ensuring that the catalyst is efficient and economical.

[0036] In this embodiment, in step S6, the AI ​​intelligent control unit periodically collects effluent water quality data through the online effluent water quality detection unit, compares the effluent pollutant concentration detection results with the effluent pollutant concentration predicted by the ozone dosage prediction model, and automatically triggers the model parameter fine-tuning program when the deviation continues to exceed the set threshold to optimize and update the parameters of the ozone dosage prediction model, thereby realizing online feedback learning of the prediction model.

[0037] In one possible embodiment, the operating cycle and deviation threshold for feedback learning are first set in the AI ​​intelligent control unit. In this embodiment, the effluent data acquisition cycle is set to 4 hours, and the deviation threshold is preset according to the effluent water quality compliance requirements. Then, during system operation, the AI ​​intelligent control unit automatically collects the pollutant concentration detection results at the effluent end according to the set 4-hour cycle, i.e., the measured value of effluent COD. The actual effluent pollutant concentration detection results are then compared with the predicted effluent results of the ozone dosage prediction model for the corresponding time period, and the deviation value between the two is calculated. Finally, the deviation changes are continuously monitored. When the deviation value exceeds the preset threshold multiple times consecutively, the system automatically triggers the model parameter fine-tuning program, optimizes and updates the parameters of the ozone dosage prediction model based on the latest accumulated operating data, and completes one online feedback learning cycle. Subsequently, the model continues to run with the updated parameters, and the above feedback process of data acquisition, deviation comparison, and parameter optimization is continuously cyclical.

[0038] Through the model self-fine-tuning mechanism triggered by periodic effluent data feedback and deviation, the predictive model achieves online self-learning and continuous optimization. This enables the system to adapt to long-term changes in influent water quality and the slow decay of catalyst activity, maintaining optimal dosing control accuracy over the long term, avoiding a decline in model performance over operating time, and ensuring long-term stable energy-saving effects of the system.

[0039] Example 2:

[0040] Reference Figure 2 This application also provides an ozone treatment device, comprising: a customized ozone oxidation tower with a height-to-diameter ratio of 5-15:1, a microporous aeration disc 5 at the bottom of the tower, an ozone catalyst packing layer 4 filled in the middle with guide plates between the layers, a Pall ring packing layer 3 at the top, a tail gas destruction and collection device 1 at the top, and multi-stage sampling ports and observation windows on the side walls; an ozone generator connected to the air inlet at the bottom of the customized ozone oxidation tower for supplying ozone into the tower; an ozone concentration detection unit installed on the air inlet pipe between the ozone generator and the customized ozone oxidation tower for monitoring the ozone dosage concentration; and an online detection unit for influent pollutant concentration installed on the system inlet pipe 8 for monitoring the ozone concentration. The system includes a real-time influent water quality monitoring unit; an effluent water quality online monitoring unit, installed on the system's effluent pipe 9, for real-time influent water quality data collection; an AI intelligent control unit, whose signal input terminals are connected to the influent pollutant concentration online detection unit and the ozone concentration detection unit, respectively, and whose control output terminal is connected to the ozone generator; the AI ​​intelligent control unit has a built-in pre-trained ozone dosage prediction model, used to calculate the instantaneous optimal ozone dosage based on real-time influent water quality data, and output control signals to the ozone generator; the AI ​​intelligent control unit also has a feedback learning module, used to automatically optimize and update the parameters of the prediction model based on the effluent water quality detection results.

[0041] In one possible embodiment, the main reaction structure of the device is first constructed: a custom-made ozone oxidation tower is manufactured, with the height-to-diameter ratio controlled to be 5-15:1. Microporous aeration discs 5 are installed at the bottom of the tower, an ozone catalyst packing layer 4 is filled in the middle of the tower with guide plates between the layers, a Pall ring packing layer 3 is installed at the top of the tower, and a tail gas destruction and collection device 1 is installed at the top of the tower. Multi-stage sampling ports and observation windows are installed on the side walls of the tower. Then, the gas and water systems are laid out: the outlet of the ozone generator is sealed to the inlet at the bottom of the custom-made ozone oxidation tower via an inlet pipe to continuously supply ozone into the tower; an ozone concentration detection unit is installed in series on the inlet pipe between the ozone generator and the custom-made ozone oxidation tower to monitor the ozone dosage concentration in real time; an online detection unit for influent pollutant concentration is installed on the inlet water pipe 8 of the system to collect influent water quality data in real time; and an ozone concentration detection unit is installed on the outlet water pipe 9 of the system. An online effluent water quality monitoring unit is used to collect effluent water quality data in real time. An electrical control loop is then established: the signal input terminal of the AI ​​intelligent control unit is connected to the signal output terminals of the online influent pollutant concentration detection unit and the ozone concentration detection unit via control cables. The control output terminal of the AI ​​intelligent control unit is connected to the control input terminal of the ozone generator. A pre-trained ozone dosage prediction model is pre-installed in the AI ​​intelligent control unit to calculate the instantaneous optimal ozone dosage based on real-time influent water quality data and output the corresponding control signal to the ozone generator. Simultaneously, a feedback learning module is integrated into the AI ​​intelligent control unit to receive effluent water quality detection results and automatically optimize and update the prediction model parameters based on the detection results. Finally, after completing all pipeline and line connections, system linkage debugging is performed to ensure normal communication between units and smooth operation logic, enabling fully automatic closed-loop operation.

[0042] By integrating a customized oxidation tower, ozone generation unit, water quality detection unit, and AI intelligent control unit according to a specific connection relationship, a complete closed-loop treatment device of detection-prediction-control-reaction-feedback is formed. This achieves a deep integration of reactor structure, catalytic technology, and intelligent control. The modules work together to improve the overall treatment efficiency and operating economy of the ozone oxidation system.

[0043] In this embodiment, the customized ozone oxidation tower further includes a circulation pump and a booster pump; the inlet and outlet of the circulation pump are connected to the lower and upper parts of the tower body respectively through circulation pipe 7, for circulating and mixing the sewage in the tower; the booster pump is connected to the outlet at the bottom of the tower body, for discharging the treated water from the bottom of the tower.

[0044] In one possible embodiment, a circulation pump and a booster pump are first installed on the customized ozone oxidation tower. The inlet of the circulation pump is connected to the water intake at the bottom of the tower through a circulation pipe 7, and the outlet of the circulation pump is connected to the water distribution interface at the top of the tower through a circulation pipe 7, forming a closed internal circulation pipeline. The inlet of the booster pump is connected to the outlet at the bottom of the tower through a pipe for outputting treated clean water. When the system starts up, the circulation pump is first turned on to transport the water at the bottom of the tower to the top of the tower through the circulation pipe 7. The water is then evenly distributed back into the tower by the water distributor 2, so that the wastewater in the tower continues to circulate, enhancing the contact and mixing effect between the wastewater and ozone and catalyst, eliminating dead zones in the tower, and improving mass transfer and reaction efficiency. After the wastewater has circulated multiple times in the tower, completed the catalytic oxidation reaction, and the water quality meets the treatment requirements, the booster pump is then turned on to pressurize the treated water at the bottom of the tower and discharge it from the system, sending it to the subsequent treatment unit or directly discharging it after it meets the standards.

[0045] By setting up a circulating pump to achieve the circulation and mixing of water in the tower, the contact effect of the gas, liquid and solid phases is further enhanced, short-circuiting and reaction dead zones in the tower are avoided, and the uniformity and treatment effect of the catalytic oxidation reaction are improved. The supporting booster pump achieves stable delivery of water from the bottom of the tower, ensuring that the system's water flow rate is controllable and the operation is continuous.

[0046] In this embodiment, the ozone catalyst in the ozone catalyst packing layer 4 is a honeycomb ceramic catalyst supported on MnO2-CeO2; the microporous aeration disc 5 is a titanium alloy microporous aeration disc 5.

[0047] In one possible embodiment, firstly, a microporous aeration disc 5 is fabricated using titanium alloy. Utilizing the excellent ozone corrosion resistance and precise microporous structure of titanium alloy, ozone passing through the aeration disc forms uniform micron-sized microbubbles. Then, a honeycomb ceramic catalyst supported on MnO2-CeO2 is prepared: using honeycomb ceramic as a carrier, soluble manganese and cerium salts are loaded onto the carrier surface and into the pores using an equal-volume impregnation method. After drying, the catalyst is placed in a muffle furnace for high-temperature calcination, converting the metal salts into MnO2-CeO2 composite active components to obtain the finished catalyst. Finally, the titanium alloy microporous aeration disc is... 5 is installed at the bottom of the customized ozone oxidation tower and is sealed to the ozone inlet pipe 6 to ensure that all ozone enters the tower body after being dispersed by the aeration discs; the honeycomb ceramic catalyst loaded with MnO2-CeO2 is neatly packed into the packing support structure in the middle of the tower body to form a catalyst packing layer; finally, after the ozone is introduced from the inlet pipe, it is dispersed into fine bubbles by the titanium alloy microporous aeration disc 5 and flows upward. When it flows through the honeycomb ceramic catalyst layer, it is rapidly decomposed under the catalytic action of the MnO2-CeO2 active sites to generate a large number of hydroxyl radicals, which efficiently and deeply oxidize the recalcitrant organic matter in the wastewater.

[0048] By adopting a titanium alloy microporous aeration disc 5, the uniformity of gas distribution and the effect of bubble refinement are ensured, and excellent ozone corrosion resistance is also provided, which can extend the service life of the equipment. The honeycomb ceramic catalyst loaded with MnO2-CeO2 combines the low pressure drop advantage of structured packing with the high catalytic activity of composite metal oxides, further improving the reaction efficiency and ozone utilization rate of catalytic ozone oxidation.

[0049] The optional embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above embodiments. Within the scope of the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the protection scope of the embodiments of the present invention.

[0050] It should also be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not describe the various possible combinations separately.

[0051] Furthermore, various different implementations of the present invention can be combined arbitrarily, as long as they do not violate the spirit of the present invention, they should also be regarded as the content disclosed in the present invention.

Claims

1. A method for operating an ozone treatment device, characterized in that, include: S1. System initialization: Load the pre-trained ozone dosage prediction model in the AI ​​intelligent control unit; S2. Data Acquisition: Real-time influent water quality data is acquired through the online detection unit for influent pollutant concentration, and the influent water quality data is transmitted to the AI ​​intelligent control unit. S3. Intelligent Prediction and Control: The AI ​​intelligent control unit calculates the instantaneous optimal ozone dosage based on real-time influent water quality data, current process parameters, and target effluent requirements using the ozone dosage prediction model, and generates a control signal. S4. Ozone dosing: The control signal drives the ozone generator to introduce precisely measured ozone into the bottom of the customized ozone oxidation tower. S5. Catalytic oxidation reaction: Ozone is distributed in the customized ozone oxidation tower to form microbubbles, which flow upward from the bottom of the tower and come into counter-current contact with the sewage that is distributed downward from the top of the tower. Under the action of the ozone catalyst in the tower, a catalytic oxidation reaction occurs. The treated water is discharged from the outlet at the bottom of the tower, and the tail gas is discharged into the atmosphere after being treated by the tail gas destruction and collection device at the top. S6. Feedback Self-Correction: The online effluent water quality detection unit periodically collects effluent water quality data and feeds it back to the AI ​​intelligent control unit. The effluent water quality detection results are compared with the model prediction results. When the deviation continues to exceed the set threshold, the parameters of the ozone dosage prediction model are automatically optimized and updated.

2. The method for operating the ozone treatment device according to claim 1, characterized in that, The customized ozone oxidation tower has a height-to-diameter ratio of 5 to 15:1; the bottom of the tower is equipped with a microporous aeration disc, the middle is filled with an ozone catalyst packing layer, and each ozone catalyst packing layer is equipped with a guide plate; the upper part of the tower is equipped with a Pall ring packing layer, which is located between the ozone catalyst packing layer and the water distributor at the top of the tower; the top of the tower is equipped with a tail gas destruction and collection device; the side wall of the tower is equipped with multi-stage sampling ports and observation windows.

3. The method for operating the ozone treatment device according to claim 1, characterized in that, In step S2, the online detection unit for pollutant concentration in the influent is installed on the influent pipe of the system to detect the concentration of organic matter in the influent in real time. The organic matter concentration indicators include at least one of COD value, UV254 value, and TOC value. In step S4, an ozone concentration detection unit is installed on the air inlet pipe of the customized ozone oxidation tower to monitor the ozone dosage concentration.

4. The method for operating the ozone treatment device according to claim 1, characterized in that, The ozone dosage prediction model built into the AI ​​intelligent control unit is a machine learning model, which includes gradient boosting decision tree, random forest or neural network model; the input features of the machine learning model include at least real-time influent COD value or real-time UV254 value, influent flow rate, pH value and water temperature, and the output is the optimal ozone dosage flow rate or optimal ozone dosage concentration.

5. The method of operating the ozone treatment device according to claim 1, characterized in that, The ozone catalyst in the ozone catalyst packing layer is a catalytic packing material that uses one of Raschig ring packing, honeycomb ceramic, porous alumina or modified activated carbon as a support and loads at least one active component of Mn, Cu, Fe, Co and Ni; the loading of the active component of the ozone catalyst is 2% to 10%.

6. The method of operating the ozone treatment device according to claim 1, characterized in that, In S6, the AI ​​intelligent control unit periodically collects effluent water quality data through the online effluent water quality detection unit, compares the effluent pollutant concentration detection results with the effluent pollutant concentration predicted by the ozone dosage prediction model, and automatically triggers the model parameter fine-tuning program when the deviation continues to exceed the set threshold to optimize and update the parameters of the ozone dosage prediction model, thereby realizing online feedback learning of the prediction model.

7. An ozone treatment device, characterized in that, A method for operating an ozone treatment apparatus as described in any one of claims 1 to 6, comprising: Customized ozone oxidation towers have a height-to-diameter ratio of 5 to 15:

1. The bottom of the tower is equipped with microporous aeration discs, the middle is filled with ozone catalyst packing layers with guide plates between the layers, the top is equipped with a Pall ring packing layer, the top is equipped with a tail gas destruction and collection device, and the side walls are equipped with multi-stage sampling ports and observation windows. An ozone generator is connected to the air inlet at the bottom of the customized ozone oxidation tower to supply ozone into the tower. An ozone concentration detection unit is installed on the air inlet pipe between the ozone generator and the customized ozone oxidation tower to monitor the ozone dosage concentration. An online detection unit for pollutant concentration in influent is installed on the influent pipe of the system to collect influent water quality data in real time. The online monitoring unit for effluent water quality is installed on the system's effluent pipe to collect effluent water quality data in real time. The AI ​​intelligent control unit has its signal input terminals connected to the online detection unit for influent pollutant concentration and the ozone concentration detection unit, respectively, and its control output terminal connected to the ozone generator. The AI ​​intelligent control unit has a built-in pre-trained ozone dosage prediction model, which is used to calculate the instantaneous optimal ozone dosage based on real-time influent water quality data and output a control signal to the ozone generator. The AI ​​intelligent control unit also has a feedback learning module, which is used to automatically optimize and update the parameters of the prediction model based on the effluent water quality detection results.

8. The ozone treatment device according to claim 7, characterized in that, The customized ozone oxidation tower also includes a circulation pump and a booster pump; the inlet and outlet of the circulation pump are connected to the lower and upper parts of the tower body respectively through circulation pipes to circulate and mix the sewage in the tower; the booster pump is connected to the outlet at the bottom of the tower body to discharge the treated water from the bottom of the tower.

9. The ozone treatment device according to claim 7, characterized in that, The ozone catalyst in the ozone catalyst packing layer is a honeycomb ceramic catalyst supported on MnO2-CeO2; the microporous aeration disc is a titanium alloy microporous aeration disc.

Images