Intelligent slurry shield waste slurry treatment system and method

Abstract

The invention discloses an intelligent slurry shield waste slurry treatment system and method, and relates to the technical field of tunnel construction, and the method comprises the following steps: collecting waste slurry data in real time, and preprocessing the waste slurry data; on the basis of the preprocessed waste slurry data, optimal waste slurry processing parameters are calculated through an AI adaptive optimization algorithm; according to the optimal treatment parameters, dynamically putting the flocculating agent through intelligent multi-point putting of the flocculating agent; in a flocculant putting process, performing state classification on each putting point by using an SVM classifier, and identifying a fault mode; and according to an identification result, dynamically triggering a multi-mode self-correction strategy to carry out fault correction. According to the method, the optimal waste slurry treatment parameters and the multi-mode self-correction strategy are dynamically calculated through the AI self-adaptive optimization algorithm, the slurry shield waste slurry treatment efficiency and reliability are greatly improved, resource waste is reduced, and the operation cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel construction, and in particular to an intelligent treatment system and method for waste slurry of a slurry shield. Background Art

[0002] With the acceleration of urbanization and the continuous advancement of infrastructure construction, shield technology is being used more and more widely in tunnel construction. However, a large amount of waste slurry containing suspended solids will be generated during slurry shield construction. If these waste slurries are not properly treated, they will cause serious pollution to the environment. Therefore, waste slurry treatment technology has gradually become an important link in slurry shield construction. Existing waste slurry treatment technologies mainly rely on the artificial control of flocculant delivery, which causes the suspended solids in the waste slurry to settle through chemical reactions. However, traditional waste slurry treatment methods are often inefficient, making it difficult to ensure the stability of waste slurry treatment effects, and lack precise control over the amount of flocculants and the energy consumption of waste slurry treatment, which can easily lead to chemical waste and increased energy consumption. In addition, existing systems usually lack intelligence and cannot adapt to complex construction environments and dynamic changes in waste slurry composition.

[0003] In the prior art, the treatment of waste slurry from slurry shields generally relies on a fixed flocculant delivery strategy, which cannot be dynamically adjusted according to real-time changes in parameters such as waste slurry flow rate and suspended solids concentration. This leads to two major problems: first, the flocculant delivery is unscientific, resulting in excessive or insufficient delivery, affecting the treatment effect and increasing operating costs; second, there is a lack of fault identification and correction mechanism during the treatment process, making it difficult to promptly detect and correct abnormal conditions at the delivery point, which can easily lead to unstable treatment effects or equipment failures. In view of these shortcomings, there is an urgent need for a waste slurry treatment system that can be intelligently and dynamically regulated based on real-time data of waste slurry, which can not only improve the efficiency of waste slurry treatment, but also promptly perform self-correction when a fault occurs, thereby reducing unnecessary chemical input and energy consumption. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a slurry shield waste slurry intelligent treatment system and method to solve the problems of inaccurate flocculant addition and lack of fault identification and correction mechanism in the prior art.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: In the first aspect, the present invention provides an intelligent treatment method for waste slurry of a slurry shield, which includes real-time collection of waste slurry data and preprocessing of the waste slurry data; based on the preprocessed waste slurry data, calculating the optimal waste slurry treatment parameters through an AI adaptive optimization algorithm; according to the optimal treatment parameters, dynamically delivering flocculants through multi-point delivery of intelligent flocculants; during the flocculant delivery process, using an SVM classifier to classify the status of each delivery point and identify fault modes; based on the identification results, dynamically triggering a multi-mode self-correction strategy to perform fault correction.

[0007] As a preferred solution of the intelligent treatment method of waste slurry of the slurry shield of the present invention, wherein: the waste slurry data includes waste slurry flow rate, turbidity, pH value, temperature and suspended solids concentration; The preprocessing of the waste pulp data is specifically as follows: Abnormal data caused by sensor failures are eliminated through data anomaly detection algorithms; Real-time data is smoothed by the sliding average method; The waste pulp data were converted into a unified dimension through normalization.

[0008] As a preferred solution of the intelligent treatment method of waste slurry of slurry shield described in the present invention, wherein: based on the pre-treated waste slurry data, the optimal waste slurry treatment parameters are calculated by AI adaptive optimization algorithm, and the specific steps are as follows: Based on the nonlinear coupling relationship between the clarity of the mud-water separation effect, the amount of flocculant used and the energy consumption in the historical data, the minimization objective function is defined to predict the comprehensive cost of waste slurry treatment. The expression is: in, Represents the comprehensive cost of waste pulp treatment, Indicates the clarity of waste pulp after treatment. Indicates the energy consumed in the waste pulp treatment process, is the maximum energy consumption allowed in the waste pulp treatment process, Indicates the maximum amount of flocculant allowed in the waste pulp treatment process. Indicates the amount of flocculant used per unit volume of waste slurry; Construct action space based on waste slurry pumping speed and flocculant dosage ; Define the state space based on the pre-treated waste slurry data and the flocculant usage in the historical data ; Comprehensive cost based on waste pulp treatment , define the reward function ; The DRL model is introduced and the parameters of the DRL model are initialized. At the same time, the waste slurry pumping speed and the amount of flocculant added are randomly initialized. Based on action space and state space And the reward function , and combined with the DRL model to form an AI adaptive optimization algorithm; At each time step t, according to the current state Select an action After executing this action, enter the next state And calculate the immediate reward through the reward function R , while rewarding For cumulative update, the expression is: Among them, α is the learning rate, is the discount factor, represents the expected cumulative reward before the update, represents the updated expected cumulative reward, Indicates the next state Next action The expected cumulative reward when , A represents the action space, represents the selection of the maximum expected cumulative reward from the action space A, is the cumulative reward function; AI adaptive optimization algorithm accumulates rewards by continuously updating , gradually select the optimal waste pulp treatment parameters under different conditions, the expression is: in, Indicates the current state The best action under The optimal action refers to the optimal waste slurry treatment parameters, including the optimal waste slurry pumping speed and the optimal flocculant dosage.

[0009] As a preferred solution of the intelligent treatment method for waste slurry of slurry shield tunneling of the present invention, wherein: the flocculant is dynamically delivered through multi-point delivery of the intelligent flocculant according to the optimal treatment parameters, and the specific steps are as follows: According to the waste slurry flow rate, use data visualization tools to draw the waste slurry flow rate distribution map in different pipeline sections to identify the changes in waste slurry flow rate in different areas; Based on the concentration of suspended solids, plot the solids concentration versus pipe position to identify the concentration changes in the area; Plan the flocculant placement points based on the changes in waste slurry flow rate and suspended solids concentration in different areas and the pipeline structure; After the placement point planning is completed, the waste slurry is treated according to the optimal waste slurry pumping speed and the optimal flocculant dosage; Based on the statistical analysis of waste slurry flow rate in historical data, define the standard threshold of waste slurry flow rate , and formulate independent flocculant delivery valve control mechanism at each delivery point; If the waste pulp flow rate at the current delivery point is monitored to be ≥ the waste pulp flow rate standard threshold When the flocculant delivery valve opening area at the current delivery point is increased, if the waste slurry flow rate at the current delivery point is less than the waste slurry flow rate standard threshold When the flocculant delivery valve opening area at the current delivery point is reduced.

[0010] As a preferred solution of the intelligent treatment method of waste slurry of the slurry shield of the present invention, in which: during the flocculant delivery process, the SVM classifier is used to classify the status of each delivery point and identify the fault type. The specific steps are as follows: During the flocculant addition process, the waste pulp data is monitored in real time, and waste pulp treatment characteristics are extracted from the waste pulp data through data analysis algorithms; Based on historical fault data, the SVM classifier is trained, and the waste pulp treatment characteristics are input into the trained SVM classifier to classify the waste pulp treatment status of different delivery points into abnormal status and normal status; Based on historical waste pulp processing data, a fault type library containing various fault types is established, and the delivery points classified as abnormal states are matched with the fault type library to identify the fault type; The various types of failures include abnormal flocculant dosing, abnormal waste slurry pumping speed, abnormal suspended solids concentration and equipment failure.

[0011] As a preferred solution of the intelligent treatment method of waste slurry of slurry shield described in the present invention, wherein: according to the identification result, the multi-mode self-correction strategy is dynamically triggered to correct the fault. The specific steps are as follows: Based on the historical waste pulp treatment data and the fault type library, a multi-mode fault correction library is established, and a three-level correction mode is formulated for the multi-mode fault correction library, and a correction strategy corresponding to the fault type library is formulated in each level of correction mode; When an abnormal state occurs at a certain delivery point, the multi-mode fault correction library is input according to the identified fault type, and the local correction strategy of the first-level correction mode corresponding to the fault type is matched to perform local correction on the current delivery point; After the local correction strategy is completed, the SVM classifier reclassifies the waste pulp treatment status of the current delivery point. If the classification result is an abnormal state, the global correction strategy of the secondary correction mode corresponding to the fault type is matched; After the global correction strategy is completed, the SVM classifier reclassifies the waste pulp treatment status of the current delivery point. If the classification result is an abnormal state, the emergency correction and alarm mechanism of the three-level correction mode corresponding to the fault type is matched.

[0012] As a preferred solution of the intelligent treatment method of slurry waste from slurry shield of the present invention, the local correction strategy of the first-level correction mode is as follows: According to the fault type of the placement point, the local waste slurry pumping speed and flocculant dosage of a single placement point are corrected; The global correction strategy of the secondary correction mode is as follows: Genetic algorithm is used to simulate the process of natural selection, generate multiple combination strategies of flocculant dosage and waste slurry pumping speed, and gradually select the global optimal correction strategy; Based on the selected optimal correction strategy, the flocculant dosage at different points is coordinated according to the changes in the waste slurry flow rate and concentration in the upstream and downstream pipelines, while the upstream waste slurry pumping speed is accelerated; The emergency correction and alarm mechanism of the three-level correction mode is as follows: After entering the third-level correction mode, it actively determines whether there is a sensor failure and flocculant supply interruption at the current delivery point; The sensor data of the delivery node in normal state is processed using waste pulp, replacing the current sensor data and triggering an alarm immediately.

[0013] In the second aspect, the present invention provides an intelligent treatment system for waste slurry of a slurry shield, including a data acquisition module, a parameter calculation module, a flocculant delivery module, a fault identification module and a fault correction module; the data acquisition module is used to collect waste slurry data in real time and pre-process the waste slurry data; the parameter calculation module is used to calculate the optimal waste slurry treatment parameters through an AI adaptive optimization algorithm based on the pre-processed waste slurry data; the flocculant delivery module is used to dynamically deliver flocculants through intelligent flocculant multi-point delivery according to the optimal treatment parameters; the fault identification module is used to use an SVM classifier to classify the status of each delivery point and identify the fault mode during the flocculant delivery process; the fault correction module is used to dynamically trigger a multi-mode self-correction strategy to correct the fault according to the identification result.

[0014] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the intelligent treatment method for slurry waste from a mud-water shield as described in the first aspect of the present invention is implemented.

[0015] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the intelligent treatment method for waste slurry of a slurry shield as described in the first aspect of the present invention is implemented.

[0016] The beneficial effects of the present invention are as follows: by dynamically calculating the optimal waste slurry treatment parameters through the AI ​​adaptive optimization algorithm, intelligent and accurate flocculant delivery is achieved, the waste or insufficiency caused by the traditional fixed delivery strategy is avoided, and the processing cost and energy consumption are significantly reduced. At the same time, the status of each delivery point is monitored in real time through the SVM classifier, the fault is quickly identified, and the dynamic correction of the fault is achieved through the multi-mode self-correction strategy to ensure timely restoration of stability under abnormal conditions. The present invention greatly improves the efficiency and reliability of slurry treatment of slurry shield, reduces resource waste, reduces operating costs, and ensures the continuity and efficiency of the treatment process. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0018] Figure 1 This is a flow chart of the intelligent treatment method for slurry waste from mud-water shield in Example 1.

[0019] Figure 2 This is a schematic diagram of the intelligent treatment system for slurry waste from a mud-water shield in Example 1. DETAILED DESCRIPTION

[0020] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0021] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0022] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0023] Example 1, reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, and provides an intelligent treatment method for waste slurry of a slurry shield, comprising the following steps: S1: Collect waste pulp data in real time and pre-process the waste pulp data.

[0024] S1.1: Slurry data include slurry flow rate, turbidity, pH value, temperature and suspended solids concentration; Furthermore, the above waste pulp data can reflect the physical and chemical properties of the waste pulp and its status during the treatment process, including fluidity, particle concentration, transparency, pH and temperature changes, thereby helping to optimize the treatment parameters.

[0025] S1.2: Preprocessing of waste pulp data is as follows: S1.2.1: Eliminate abnormal data caused by sensor failure through data anomaly detection algorithm; For example, when a sensor continuously reports a waste slurry flow rate of zero or an extreme abnormal value (such as high temperature or low turbidity outside a reasonable range), the data can be determined to be abnormal and discarded.

[0026] S1.2.2: Smooth the real-time data by sliding average method; For example, the continuous data within a period of time is averaged, such as calculating the average value once every 5 consecutive flow rate data, to reduce the impact of short-term fluctuations and obtain a smoother numerical curve.

[0027] S1.2.3: Convert waste pulp data to a uniform dimension by normalization.

[0028] For example, waste slurry data of different units (such as flow rate, suspended solids concentration, pH value, etc.) are scaled to the same range, so that the algorithm can be uniformly processed and compared.

[0029] S2: Based on the pre-treated waste pulp data, the AI ​​adaptive optimization algorithm is used to calculate the optimal waste pulp treatment parameters.

[0030] S2.1: Based on the nonlinear coupling relationship between the clarity of the mud-water separation effect, the amount of flocculant used and the energy consumption in the historical data, the minimization objective function is defined to predict the comprehensive cost of waste slurry treatment. The expression is: in, Represents the comprehensive cost of waste pulp treatment, Indicates the clarity of waste pulp after treatment. Indicates the energy consumed in the waste pulp treatment process, is the maximum energy consumption allowed in the waste pulp treatment process, Indicates the maximum amount of flocculant allowed in the waste pulp treatment process. Indicates the amount of flocculant used per unit volume of waste slurry; Furthermore, in combination with key variables in waste pulp treatment, such as clarity , Flocculant usage and energy consumption , and its maximum allowable value, accurately describe the nonlinear coupling relationship between them. This formula quantifies the cost changes under different treatment conditions through the combination of multiple terms, thus providing a theoretical basis for optimizing the amount of flocculant and energy consumption, ensuring that the waste pulp treatment cost is minimized while meeting the clarification effect.

[0031] S2.2: Constructing action space based on waste slurry pumping speed and flocculant dosage ; It should be noted that the action space A includes all possible combinations of waste pulp pumping speed and flocculant dosage, and is used to represent the operational options that can be performed under different states.

[0032] S2.3: Define the state space based on the pre-treated waste slurry data and the flocculant usage in the historical data ; It should be noted that the state space It represents the combination of the pre-treated waste slurry characteristic data (such as flow rate, suspended solid concentration, etc.) and the amount of flocculant used at each moment, reflecting the current state of the waste slurry treatment system.

[0033] S2.4: Comprehensive cost based on waste pulp treatment , define the reward function ; It should be noted that the better the processing effect (that is, the smaller the objective function J), the greater the reward The larger the value, the worse the treatment effect (i.e. the larger the objective function J), and the higher the reward The smaller.

[0034] S2.5: introduce the DRL model, initialize the parameters of the DRL model, and randomly initialize the waste slurry pumping speed and flocculant dosage; It should be noted that the DRL model is introduced to automatically explore and optimize the pumping speed and flocculant dosage in the waste slurry treatment process through deep reinforcement learning, so as to find the optimal treatment strategy under different conditions.

[0035] S2.6: Based on action space, state space S and reward function , and combined with the DRL model to form an AI adaptive optimization algorithm; At each time step t, according to the current state Select an action After executing this action, enter the next state And calculate the immediate reward through the reward function R , while rewarding For cumulative update, the expression is: Among them, α is the learning rate, is the discount factor, represents the expected cumulative reward before the update, represents the updated expected cumulative reward, Indicates the next state Next action The expected cumulative reward when , A represents the action space, represents the selection of the maximum expected cumulative reward from the action space A, is the cumulative reward function; It should be noted that Indicates the current state Next action The expected cumulative reward at that time is to update the reward estimate for that state and action; Indicates that an action is being performed Afterwards, based on the immediate reward and the next state The maximum expected reward of is the updated cumulative reward estimate, which is the new estimate after learning.

[0036] This expression gradually optimizes the actions selected in different states by continuously updating the reward value of the state-action pair.

[0037] S2.7: AI adaptive optimization algorithm accumulates rewards by continuously updating , gradually select the optimal waste pulp treatment parameters under different conditions, the expression is: in, Indicates the current state The best action under Optimal action refers to the optimal waste slurry treatment parameters, including the optimal waste slurry pumping speed and the optimal flocculant dosage.

[0038] It should also be noted that the AI ​​adaptive optimization algorithm firstly Select an action (such as a specific amount of flocculant and pumping speed), after executing this action, enter the next state , and based on the immediate reward The Q value is updated, so that at each time step t, the algorithm gradually adjusts the action selection by continuously comparing the cumulative rewards of the current action and possible future actions, and finally finds the optimal parameter combination that can minimize the waste pulp treatment cost under different states.

[0039] S3: Dynamically deliver flocculants through multi-point delivery of intelligent flocculants according to the optimal treatment parameters.

[0040] S3.1: Based on the waste slurry flow rate, use data visualization tools to draw the waste slurry flow rate distribution map in different pipeline sections and identify the changes in waste slurry flow rate in different areas; Based on the suspended solids concentration, draw a curve chart showing the change of suspended solids concentration with the pipeline position to identify the change of suspended solids concentration in the current area; It should be noted that by drawing the flow velocity distribution diagram of the waste slurry in different pipeline sections and the curve diagram of the suspended solids concentration changing with the position, the dynamic changes of the waste slurry flow and particle distribution can be intuitively displayed, which helps to identify areas with higher flow velocity or higher suspended solids concentration, thereby providing data support for optimizing the flocculant delivery point and delivery amount, ensuring that the flocculant can be effectively mixed in the key area and achieve the best flocculation effect.

[0041] S3.2: Plan the flocculant placement points based on the changes in waste slurry flow rate and suspended solids concentration in different areas and the pipeline structure; The planned flocculant placement points are as follows: Principles of placement point layout planning: Priority placement in high flow rate areas: In areas with high waste slurry flow rate, the number of placement points should be appropriately increased to ensure that the flocculant can be effectively mixed with the waste slurry and avoid the flocculant being washed away prematurely.

[0042] Key placement points in high-concentration areas: In areas with high suspended solids concentration, increase the density of placement points so that there are enough flocculants in contact with the solid particles in the waste slurry to promote the flocculation effect.

[0043] Distribution of points in bends and narrow sections: When the waste slurry flows through the bends or narrow sections of the pipeline, solid particles are easily deposited. It is recommended to arrange delivery points in these areas to avoid excessive local deposition that affects the treatment effect.

[0044] Number and spacing of placement points: Match the density of the delivery points according to the waste slurry flow rate and solid concentration: in high flow rate and high concentration areas, the spacing between delivery points should be appropriately shortened to ensure that the flocculant can be evenly distributed. In low flow rate and low concentration areas, the delivery points can be relatively sparse.

[0045] Spacing optimization: By analyzing the residence time and flow characteristics of the waste slurry in different areas, the diffusion range of the flocculant is estimated, and the distance between the delivery points is reasonably determined to ensure that the flocculant is fully diffused in the waste slurry without excessive waste.

[0046] S3.3: After the placement point planning is completed, the waste slurry is treated according to the optimal waste slurry pumping speed and the optimal flocculant dosage; It should be noted that the waste slurry is pumped to the treatment system through a pipeline, and during the flow process, flocculants are dynamically injected into the waste slurry. The flocculants react with the suspended solid particles in the waste slurry, causing the particles to combine to form larger flocs. As the waste slurry flows into the sedimentation tank, these flocs settle to the bottom of the tank due to gravity, forming a solid layer, and the clarified water above is gradually separated. The treated clarified water can be discharged or further purified, while the deposited solids are regularly cleaned or subsequently treated. Throughout the process, the amount of flocculant added and the pumping speed are dynamically adjusted based on the real-time monitored flow rate, suspended solids concentration and other data to ensure the best solid-liquid separation effect.

[0047] S3.4: Define the standard threshold of waste slurry flow rate based on the statistical analysis of waste slurry flow rate in historical data , and formulate independent flocculant delivery valve control mechanism at each delivery point; If the waste pulp flow rate at the current delivery point is monitored to be ≥ the waste pulp flow rate standard threshold When the flocculant delivery valve opening area at the current delivery point is increased, if the waste slurry flow rate at the current delivery point is less than the waste slurry flow rate standard threshold When the flocculant delivery valve opening area at the current delivery point is reduced.

[0048] It should be noted that by setting the standard threshold of waste pulp flow rate , and dynamically adjust the opening area of ​​the flocculant delivery valve according to the real-time monitored flow rate, which can ensure that the flocculant delivery amount is increased when the flow rate is high to enhance the flocculation effect, and the delivery amount is reduced when the flow rate is low to avoid waste, thereby achieving precise control of flocculant delivery and optimizing waste slurry treatment efficiency and cost.

[0049] S4: During the flocculant delivery process, the SVM classifier is used to classify the status of each delivery point and identify the failure mode.

[0050] S4.1: During the flocculant addition process, the waste pulp data is monitored in real time, and waste pulp treatment characteristics are extracted from the waste pulp data through data analysis algorithms; Furthermore, the waste slurry treatment characteristics include, settling rate, solid-liquid separation efficiency, floc formation quality and energy efficiency; S4.2: Based on the historical fault data, the SVM classifier is trained, the waste pulp treatment characteristics are input into the trained SVM classifier, and the waste pulp treatment status of different delivery points is divided into abnormal status and normal status; It should be noted that the SVM classifier has a strong ability in processing high-dimensional data and nonlinear classification problems. It can effectively distinguish between normal and abnormal states in waste pulp treatment by constructing a hyperplane. Especially when there is less or unbalanced fault data, the SVM classifier can still maintain a high classification accuracy, ensuring that the waste pulp treatment system can identify potential anomalies in real time and make timely adjustments.

[0051] S4.3: Based on the historical waste pulp treatment data, a fault type library containing multiple fault types is established, and the delivery points classified as abnormal states are matched with the fault type library to identify the fault type; It should be noted that based on the actual fault events that occurred during the historical waste slurry treatment process, different fault types were summarized and the fault labels of each historical data were marked. Finally, these marked fault data were classified and sorted to build a library containing multiple fault types, and the characteristic patterns of each fault were refined, such as abnormal flow rate fluctuations or changes in suspended solids concentration, for subsequent fault type identification.

[0052] The various types of failures include abnormal flocculant dosing, abnormal waste slurry pumping speed, abnormal suspended solids concentration and equipment failure.

[0053] S5: Based on the identification results, dynamically trigger the multi-mode self-correction strategy to perform fault correction.

[0054] S5.1: Based on the historical waste pulp treatment data and the fault type library, a multi-mode fault correction library is established, and a three-level correction mode is formulated for the multi-mode fault correction library, and a correction strategy corresponding to the fault type library is formulated in each level of correction mode; The specific process of establishing a multi-mode fault correction library is as follows: extract different fault types and their corresponding characteristic patterns from the historical waste pulp treatment data and fault type library, such as abnormal flocculant injection or unstable pump speed, etc. Then, based on these fault types, analyze the impact range and severity of each fault, and formulate a correction strategy for each fault.

[0055] It should also be noted that the first-level correction mode is local correction, the second-level correction mode is global correction, and the third-level correction mode is an emergency correction and alarm mechanism.

[0056] S5.1.1: When an abnormal state occurs at a certain delivery point, the multi-mode fault correction library is input according to the identified fault type, and the local correction strategy of the first-level correction mode corresponding to the fault type is matched to perform local correction on the current delivery point; S5.1.2: After the local correction strategy is completed, the SVM classifier reclassifies the waste pulp treatment status of the current delivery point. If the classification result is an abnormal state, the global correction strategy of the secondary correction mode corresponding to the fault type is matched; S5.1.3: After the global correction strategy is completed, the SVM classifier reclassifies the waste pulp treatment status of the current delivery point. If the classification result is an abnormal state, the emergency correction and alarm mechanism of the three-level correction mode corresponding to the fault type is matched.

[0057] It should be noted that through the multi-level correction mode, the correction strength is gradually increased to ensure that the waste pulp treatment system can be accurately and effectively adjusted under different fault severity levels. Specifically: through the local correction strategy, small-scale anomalies can be quickly resolved to avoid over-adjustment; if the problem is not resolved, the entire system is optimized on a larger scale through the global correction strategy; if the fault still cannot be eliminated, the emergency correction and alarm mechanism is activated. This hierarchical correction process can effectively improve stability, reduce downtime, and minimize the impact of faults on waste pulp treatment efficiency and costs.

[0058] S5.2: The local correction strategy of the first-level correction mode is as follows: According to the fault type of the placement point, the local waste slurry pumping speed and flocculant dosage of a single placement point are corrected; Furthermore, if the waste slurry flow rate is too high, the pumping speed will be appropriately reduced to ensure that the flocculant has enough time to fully mix with the waste slurry; if the amount of flocculant added is too much or too little, the amount of flocculant added will be optimized according to real-time data to ensure that the amount of flocculant used matches the suspended solids concentration of the waste slurry. The entire correction process is to restore the normal waste slurry treatment state through small, local adjustments to avoid affecting the overall operating efficiency.

[0059] S5.3: The global correction strategy of the secondary correction mode is as follows: Genetic algorithm is used to simulate the process of natural selection, generate multiple combination strategies of flocculant dosage and waste slurry pumping speed, and gradually select the global optimal correction strategy; Based on the selected optimal correction strategy, the flocculant dosage at different points is coordinated according to the changes in the waste slurry flow rate and concentration in the upstream and downstream pipelines, while the upstream waste slurry pumping speed is accelerated; Furthermore, if the flow rate in the upstream pipeline is too low and the concentration is high, the amount of flocculant added should be appropriately reduced to prevent excessive flocculation from causing floc accumulation in the downstream pipeline. At the same time, by speeding up the upstream waste slurry pumping speed, the waste slurry can be ensured to pass through the upstream pipeline quickly, reducing the risk of downstream blockage caused by slow flow rate, thereby achieving a dynamic balance between the flow and treatment efficiency between the upstream and downstream pipelines, and preventing blockage or uneven treatment in local areas.

[0060] S5.3: The emergency correction and alarm mechanism of the three-level correction mode is as follows: After entering the third-level correction mode, it actively determines whether there is a sensor failure and flocculant supply interruption at the current delivery point; The sensor data of the delivery node in normal state is processed using waste pulp, replacing the current sensor data and triggering an alarm immediately.

[0061] Furthermore, after entering the third-level correction mode, historical data analysis and real-time status monitoring are first used to proactively determine whether there are problems such as sensor failure or flocculant supply interruption at the current delivery point. If a sensor failure is detected, the sensor data of the adjacent or upstream and downstream delivery nodes where the waste pulp treatment is in normal state is automatically called as temporary replacement data for the current node to ensure that the waste pulp treatment process is not interrupted. At the same time, the alarm mechanism is immediately triggered to notify the operation and maintenance personnel to conduct on-site inspections and repairs, thereby avoiding waste pulp treatment failures or abnormalities caused by sensor failures or flocculant supply interruptions. This mechanism can not only maintain the continuity of treatment in the short term, but also respond to and repair potential hardware failures in a timely manner.

[0062] This embodiment also provides a slurry shield waste slurry intelligent treatment system, including: a data acquisition module, a parameter estimation module, a flocculant delivery module, a fault identification module and a fault correction module; Data acquisition module, used for real-time collection of waste pulp data and pre-processing of waste pulp data; Parameter estimation module, used to estimate the optimal waste pulp treatment parameters based on the pre-treated waste pulp data through AI adaptive optimization algorithm; Flocculant delivery module, used to dynamically deliver flocculants through intelligent multi-point delivery according to the optimal treatment parameters; The fault identification module is used to classify the status of each delivery point and identify the fault mode using the SVM classifier during the flocculant delivery process; The fault correction module is used to dynamically trigger a multi-mode self-correction strategy to correct the fault according to the identification result.

[0063] This embodiment also provides a computer device, which is suitable for the intelligent treatment method of slurry waste from slurry shield, including: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the intelligent treatment method of slurry waste from slurry shield as proposed in the above embodiment.

[0064] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0065] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the intelligent treatment method for waste slurry of a slurry shield proposed in the above embodiment is implemented; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0066] In summary, the present invention achieves intelligent and accurate flocculant delivery through: AI adaptive optimization algorithm dynamically calculates the optimal waste slurry treatment parameters, avoids the waste or insufficiency caused by traditional fixed delivery strategies, and significantly reduces processing costs and energy consumption. At the same time, the status of each delivery point is monitored in real time through the SVM classifier, faults are quickly identified, and dynamic correction of faults is achieved through a multi-mode self-correction strategy to ensure timely restoration of stability under abnormal circumstances. The present invention greatly improves the efficiency and reliability of slurry treatment of slurry shields, reduces resource waste, reduces operating costs, and ensures the continuity and efficiency of the treatment process.

[0067] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An intelligent method for processing waste slurry from a slurry shield, characterized by: include, Collect waste pulp data in real time and pre-process the waste pulp data; Based on the pre-treated waste pulp data, the optimal waste pulp treatment parameters are calculated through the AI ​​adaptive optimization algorithm; Dynamically place flocculants according to optimal treatment parameters through intelligent multi-point placement of flocculants; During the flocculant placement process, the SVM classifier is used to classify the status of each placement point and identify the failure mode; According to the identification results, the multi-mode self-correction strategy is dynamically triggered to perform fault correction.

2. The intelligent treatment method for waste slurry of a slurry shield as claimed in claim 1, characterized in that: The waste slurry data include waste slurry flow rate, turbidity, pH value, temperature and suspended solids concentration; The preprocessing of the waste pulp data is specifically as follows: Abnormal data caused by sensor failures are eliminated through data anomaly detection algorithms; Real-time data is smoothed by sliding average method; The waste pulp data were converted into a unified dimension through normalization.

3. The intelligent treatment method for waste slurry of a slurry shield as claimed in claim 2, characterized in that: Based on the pre-treated waste pulp data, the optimal waste pulp treatment parameters are calculated by the AI ​​adaptive optimization algorithm. The specific steps are as follows: Based on the nonlinear coupling relationship between the clarity of the mud-water separation effect, the amount of flocculant used and the energy consumption in the historical data, the minimization objective function is defined to predict the comprehensive cost of waste slurry treatment. The expression is: in, Represents the comprehensive cost of waste pulp treatment, Indicates the clarity of waste pulp after treatment. Indicates the energy consumed in the waste pulp treatment process, is the maximum energy consumption allowed in the waste pulp treatment process, Indicates the maximum amount of flocculant allowed in the waste pulp treatment process. Indicates the amount of flocculant used per unit volume of waste slurry; Construct action space based on waste slurry pumping speed and flocculant dosage ; Define the state space based on the pre-treated waste slurry data and the flocculant usage in the historical data ; Comprehensive cost based on waste pulp treatment , define the reward function ; The DRL model is introduced and the parameters of the DRL model are initialized. At the same time, the waste slurry pumping speed and the amount of flocculant added are randomly initialized. Based on action space and state space And the reward function , and combined with the DRL model to form an AI adaptive optimization algorithm; At each time step t, according to the current state space Status in Select an action After executing this action, enter the next state And calculate the immediate reward through the reward function R , while rewarding For cumulative update, the expression is: Among them, α is the learning rate, is the discount factor, represents the expected cumulative reward before the update, represents the updated expected cumulative reward, Indicates the next state Next action The expected cumulative reward when , A represents the action space, represents the selection of the maximum expected cumulative reward from the action space A, is the cumulative reward function; AI adaptive optimization algorithm accumulates rewards by continuously updating , gradually select the optimal waste pulp treatment parameters under different conditions, the expression is: in, Indicates the current state The best action under The optimal action refers to the optimal waste slurry treatment parameters, including the optimal waste slurry pumping speed and the optimal flocculant dosage.

4. The intelligent treatment method for waste slurry of a slurry shield as claimed in claim 3 is characterized by: According to the optimal treatment parameters, the flocculant is dynamically placed through multi-point placement of the intelligent flocculant. The specific steps are as follows: According to the waste slurry flow rate, use data visualization tools to draw the waste slurry flow rate distribution map in different pipeline sections to identify the changes in waste slurry flow rate in different areas; Based on the concentration of suspended solids, plot the solids concentration versus pipe position to identify the concentration changes in the area; Plan the flocculant placement points based on the changes in waste slurry flow rate and suspended solids concentration in different areas and the pipeline structure; After the placement point planning is completed, the waste slurry is treated according to the optimal waste slurry pumping speed and the optimal flocculant dosage; Based on the statistical analysis of waste slurry flow rate in historical data, define the standard threshold of waste slurry flow rate , and formulate independent flocculant delivery valve control mechanism at each delivery point; If the waste pulp flow rate at the current delivery point is monitored to be ≥ the waste pulp flow rate standard threshold When the flocculant delivery valve opening area at the current delivery point is increased, if the waste slurry flow rate at the current delivery point is less than the waste slurry flow rate standard threshold When the flocculant delivery valve opening area at the current delivery point is reduced.

5. The intelligent treatment method for waste slurry of a slurry shield as claimed in claim 4, characterized in that: In the flocculant delivery process, the SVM classifier is used to classify the status of each delivery point and identify the fault type. The specific steps are as follows: During the flocculant addition process, the waste pulp data is monitored in real time, and waste pulp treatment characteristics are extracted from the waste pulp data through data analysis algorithms; Based on historical fault data, the SVM classifier is trained, and the waste pulp treatment characteristics are input into the trained SVM classifier to classify the waste pulp treatment status of different delivery points into abnormal status and normal status; Based on historical waste pulp processing data, a fault type library containing various fault types is established, and the delivery points classified as abnormal states are matched with the fault type library to identify the fault type; The various types of failures include abnormal flocculant dosing, abnormal waste slurry pumping speed, abnormal suspended solids concentration and equipment failure.

6. The intelligent treatment method for waste slurry of a slurry shield as claimed in claim 5, characterized in that: According to the identification results, the multi-mode self-correction strategy is dynamically triggered to correct the fault. The specific steps are as follows: Based on the historical waste pulp treatment data and the fault type library, a multi-mode fault correction library is established, and a three-level correction mode is formulated for the multi-mode fault correction library, and a correction strategy corresponding to the fault type library is formulated in each level of correction mode; When an abnormal state occurs at a certain delivery point, the identified fault type is input into the multi-mode fault correction library, and the local correction strategy of the first-level correction mode corresponding to the fault type is matched to perform local correction on the current delivery point; After the local correction strategy is completed, the SVM classifier reclassifies the waste pulp treatment status of the current delivery point. If the classification result is an abnormal state, the global correction strategy of the secondary correction mode corresponding to the fault type is matched; After the global correction strategy is completed, the SVM classifier reclassifies the waste pulp treatment status of the current delivery point. If the classification result is an abnormal state, the emergency correction and alarm mechanism of the three-level correction mode corresponding to the fault type is matched.

7. The intelligent treatment method for waste slurry of a slurry shield as claimed in claim 6, characterized in that: The local correction strategy of the first-level correction mode is as follows: According to the fault type of the placement point, the local waste slurry pumping speed and flocculant dosage of a single placement point are corrected; The global correction strategy of the secondary correction mode is as follows: Genetic algorithm is used to simulate the process of natural selection, generate multiple combination strategies of flocculant dosage and waste slurry pumping speed, and gradually select the global optimal correction strategy; Based on the selected optimal correction strategy, the flocculant dosage at different points is coordinated according to the changes in the waste slurry flow rate and concentration in the upstream and downstream pipelines, while the upstream waste slurry pumping speed is accelerated; The emergency correction and alarm mechanism of the three-level correction mode is as follows: After entering the third-level correction mode, it actively determines whether there is a sensor failure and flocculant supply interruption at the current delivery point; The sensor data of the delivery node in normal state is processed using waste pulp, replacing the current sensor data and triggering an alarm immediately.

8. An intelligent treatment system for waste slurry of a slurry shield, based on the intelligent treatment method for waste slurry of a slurry shield according to any one of claims 1 to 7, characterized in that: Including data acquisition module, parameter estimation module, flocculant delivery module, fault identification module and fault correction module; Data acquisition module, used for real-time collection of waste pulp data and pre-processing of waste pulp data; Parameter estimation module, used to estimate the optimal waste pulp treatment parameters based on the pre-treated waste pulp data through AI adaptive optimization algorithm; Flocculant delivery module, used to dynamically deliver flocculants through intelligent multi-point delivery according to the optimal treatment parameters; The fault identification module is used to classify the status of each delivery point and identify the fault mode using the SVM classifier during the flocculant delivery process; The fault correction module is used to dynamically trigger a multi-mode self-correction strategy to correct the fault according to the identification result.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the intelligent treatment method of slurry waste from mud shield according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the intelligent treatment method of slurry waste from a slurry shield as described in any one of claims 1 to 7 are implemented.

Images