Intelligent solid waste environment-friendly treatment system and method

Abstract

The invention relates to the technical field of environment-friendly treatment, and particularly discloses an intelligent solid waste environment-friendly treatment system and method.According to the intelligent solid waste environment-friendly treatment system and method, municipal waste is subjected to landfill, leachate generated in the biodegradation process of the municipal waste is collected, and meanwhile key state parameters such as the pH value of the leachate and landfill gas components in the degradation process are monitored in real time; a deep learning algorithm is introduced to carry out sequential evolution mode analysis on multi-dimensional key state parameters in the degradation process, meanwhile, the current biodegradation reaction stage is self-adaptively recognized in combination with the current reacted time, and then an adaptive recharge leachate target pH value is determined based on the recognition result of the reaction stage. And adjusting the pH value of the collected leachate according to the pH value of the collected leachate, and recharging the adjusted leachate into the landfill area to dynamically adjust the pH value environment of the leachate. According to the method, the optimal biochemical environment can be provided for microbial flora in different stages through intelligent sensing and staged regulation and control of the garbage degradation process, so that the garbage degradation efficiency and the gas production rate are improved.

Description

Technical Field

[0001] This application relates to the field of environmental protection technology, and more specifically, to an intelligent solid waste environmental protection treatment system and method. Background Technology

[0002] With the acceleration of global urbanization and the continuous growth of population, the amount of urban solid waste generated is increasing daily, posing a severe challenge to the ecological environment and public health. Landfill, as a widely used end-of-life treatment method for solid waste, can centrally process large amounts of waste, but its core biodegradation process is extremely slow and uncontrolled under natural conditions, resulting in problems such as long stabilization periods, incomplete degradation, and low gasification rates.

[0003] In response, the invention patent with publication number CN109622565B proposes an environmentally friendly solid waste treatment system. This system collects leachate generated in landfill areas, adjusts the pH value of the leachate, and then re-injects the adjusted leachate back into the landfill. This utilizes the microorganisms and nutrients rich in the leachate to accelerate the decomposition rate of organic matter, thereby speeding up the degradation process of organic matter.

[0004] However, existing technologies mainly use a fixed pH target value (such as 8.0) to adjust the acidity or alkalinity of the recharge leachate, while ignoring the fact that waste biodegradation is a highly dynamic and complex process that includes continuous stages with distinct characteristics, such as aerobic decomposition, facultative anaerobic decomposition, acid production, methanogenesis, and composting. Moreover, the dominant microbial communities and their metabolic activities, as well as the required suitable environmental conditions (especially pH value) for each stage, are significantly different. This makes it easy for the adjustment strategy to be seriously mismatched with the optimal conditions required for the current actual degradation stage when using a fixed pH target to adjust the acidity or alkalinity of the leachate for recharge. This not only fails to fully stimulate microbial activity to achieve efficient degradation, but may even inhibit the function of key microbial groups (such as methanogens), thereby weakening the acceleration effect and making it difficult to achieve global optimization control of the degradation process.

[0005] Therefore, there is a need for an intelligent solid waste environmental protection treatment system and method. Summary of the Invention

[0006] To address the aforementioned technical problems, this application is proposed. Embodiments of this application provide an intelligent solid waste environmental protection treatment system and method. This system landfills municipal solid waste and collects the leachate generated during its biodegradation process. Simultaneously, it monitors key state parameters such as pH value and landfill gas composition of the leachate in real time during the biodegradation process. By introducing a deep learning algorithm, it analyzes the temporal evolution patterns of multidimensional key state parameters during the waste biodegradation process. Furthermore, it adaptively identifies the current stage of waste biodegradation reaction based on the current reaction time. Subsequently, based on the identification results of the reaction stage, it determines a suitable target pH value for the reinjected leachate and adjusts the pH of the collected leachate accordingly. The adjusted leachate is then reinjected into the landfill area to dynamically regulate its pH environment. This method, through intelligent sensing and phased control of the waste degradation process, can provide an optimal biochemical environment for microbial communities at different stages, thereby improving waste degradation efficiency and gas production rate.

[0007] Accordingly, according to one aspect of this application, an intelligent environmentally friendly method for treating solid waste is provided, comprising:

[0008] Municipal waste is landfilled in landfill areas for biodegradation, and the resulting leachate is collected;

[0009] The time series of biodegradation reaction state parameters within a predetermined time window are obtained at a predetermined sampling frequency. The biodegradation reaction state parameters include leachate pH value, CH4 concentration, CO2 concentration, H2 concentration and O2 concentration in landfill gas.

[0010] The time series of the biodegradation reaction state parameters and the current reaction time are input into the waste biodegradation reaction stage identification model to obtain the current waste biodegradation reaction stage.

[0011] Based on the current stage of waste biodegradation reaction, the target pH value of the reinjection leachate is determined;

[0012] The pH of the leachate is adjusted based on the deviation between the target pH value and the current pH value of the recharge leachate to obtain the recharge leachate, and the recharge leachate is then recharged into the landfill.

[0013] According to another aspect of this application, an intelligent solid waste environmental protection treatment system is provided, comprising:

[0014] The leachate collection module is used to collect leachate generated from municipal solid waste that has been biodegraded in landfill areas.

[0015] A biodegradation reaction status monitoring module is used to acquire the time series of biodegradation reaction status parameters within a predetermined time window at a predetermined sampling frequency. The biodegradation reaction status parameters include leachate pH value, CH4 concentration, CO2 concentration, H2 concentration and O2 concentration in landfill gas.

[0016] The biodegradation reaction stage identification module is used to input the time series of the biodegradation reaction state parameters and the current reaction time into the waste biodegradation reaction stage identification model to obtain the current waste biodegradation reaction stage.

[0017] The target pH value determination module is used to determine the target pH value of the reinjection leachate based on the current stage of waste biodegradation reaction.

[0018] A pH adjustment module is used to adjust the pH of the leachate based on the deviation between the target pH value of the recharge leachate and the current pH value of the leachate to obtain recharge leachate, and to recharge the recharge leachate back into the landfill area.

[0019] Compared with existing technologies, the intelligent solid waste environmental protection treatment system and method provided in this application involves landfilling municipal solid waste and collecting leachate generated during its biodegradation process. Simultaneously, it monitors key state parameters such as pH value and landfill gas composition of the leachate in real time during the biodegradation process. By introducing a deep learning algorithm, it analyzes the temporal evolution patterns of multidimensional key state parameters during the waste biodegradation process and adaptively identifies the current stage of waste biodegradation reaction based on the current reaction time. Subsequently, based on the identification results of the reaction stage, it determines the appropriate target pH value for the reinjected leachate and adjusts the pH of the collected leachate accordingly. The adjusted leachate is then reinjected into the landfill area to dynamically regulate its pH environment. This method, through intelligent sensing and phased control of the waste degradation process, can provide an optimal biochemical environment for microbial communities at different stages, thereby improving waste degradation efficiency and gas production rate. Attached Figure Description

[0020] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0021] Figure 1 This is a flowchart of an intelligent solid waste environmental protection treatment method according to an embodiment of this application.

[0022] Figure 2This is a flowchart of step S3 in the intelligent solid waste environmental protection treatment method according to an embodiment of this application.

[0023] Figure 3 This is a schematic diagram of the data flow in step S3 of the intelligent solid waste environmental protection treatment method according to an embodiment of this application.

[0024] Figure 4 This is a flowchart of step S32 in the intelligent solid waste environmental protection treatment method according to an embodiment of this application.

[0025] Figure 5 This is a block diagram of an intelligent solid waste environmental protection treatment system according to an embodiment of this application. Detailed Implementation

[0026] Hereinafter, exemplary embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0027] Figure 1 This is a flowchart of an intelligent solid waste environmental protection treatment method according to an embodiment of this application. Figure 1 As shown, the intelligent solid waste environmental protection treatment method according to an embodiment of this application includes the following steps: S1, landfilling municipal solid waste in a landfill area for biodegradation and collecting the generated leachate; S2, acquiring the time series of biodegradation reaction state parameters within a predetermined time window at a predetermined sampling frequency, wherein the biodegradation reaction state parameters include the pH value of the leachate, the CH4 concentration, CO2 concentration, H2 concentration, and O2 concentration in the landfill gas; S3, inputting the time series of the biodegradation reaction state parameters and the current reaction time into a waste biodegradation reaction stage identification model to obtain the current waste biodegradation reaction stage; S4, determining the target pH value of the recharge leachate based on the current waste biodegradation reaction stage; S5, adjusting the pH of the leachate based on the deviation between the target pH value of the recharge leachate and the current pH value of the leachate to obtain the recharge leachate, and recharging the recharge leachate back into the landfill area.

[0028] In the aforementioned intelligent solid waste environmental protection treatment method, step S1 involves landfilling municipal solid waste in a landfill area for biodegradation and collecting the resulting leachate. It should be understood that the environmental protection treatment of solid waste essentially involves stabilizing and rendering it harmless, and sanitary landfill is currently the basic and mainstream method for treating municipal solid waste. Therefore, this application, based on modern sanitary landfill engineering, initiates the natural biodegradation process of waste by constructing a landfill area that meets environmental standards.

[0029] In the specific implementation process, modern landfill sites will be selected and constructed in accordance with environmental protection standards. During the site selection process, factors such as geological conditions, hydrogeological conditions, climate characteristics, transportation convenience, service radius, and distance from residential areas must be comprehensively considered to ensure that the selected site has good geological stability, low groundwater level, low permeability coefficient, and is far away from drinking water sources and ecologically sensitive areas.

[0030] Once the site selection was finalized, the construction of the reservoir strictly adhered to the principles of seepage prevention, drainage, and ease of monitoring. First, the seepage prevention system is the core barrier of the landfill area, its purpose being to prevent leachate from seeping downwards and contaminating groundwater, and from spreading laterally and polluting the soil. At the bottom and slopes of the reservoir, a multi-layered composite seepage prevention structure is laid. The bottom layer is typically a compacted, low-permeability clay layer, at least 0.5 meters thick, with a permeability coefficient of less than 10⁻⁷ cm / s, forming a natural barrier. Above this, at least two layers of high-density polyethylene (HDPE) geomembrane are laid, each at least 2.0 mm thick, and all joints are ensured to be intact and tight using hot-melt welding technology, achieving an almost completely impermeable effect. Between the two HDPE membrane layers, a leakage detection layer is usually installed to monitor the integrity of the geomembrane in real time; once leakage is detected, it can be located and repaired promptly. On top of the HDPE membrane, a geotextile is laid as a protective layer to prevent mechanical damage to the geomembrane from waste. The synergistic effect of this multi-layered composite seepage prevention structure creates a relatively closed environment inside the reservoir area, ensuring that leachate does not overflow in an uncontrolled manner.

[0031] Secondly, the leachate drainage system is a crucial component complementing the impermeable layer. Despite robust impermeability, a significant amount of leachate still accumulates within the landfill. If not drained promptly, this leachate can create high water levels, increasing pressure on the impermeable layer and potentially leading to slope instability. Therefore, an efficient leachate drainage layer is constructed above the impermeable layer. This drainage layer typically consists of uniformly sized coarse sand or gravel (such as a gravel drainage layer), with a thickness of at least 0.5 meters and good permeability. Within the gravel layer, porous blind pipes (such as HDPE perforated pipes) are evenly distributed. These blind pipes are laid radially in a fishbone or dendritic pattern, converging at the lowest point of the landfill's bottom into a main collection pipe. Driven by gravity, the leachate permeates through the waste layer and the gravel drainage layer, enters the blind pipes, and ultimately collects through the main collection pipe into a leachate equalization tank located in a low-lying area outside or inside the landfill. Equalization tanks are typically constructed of reinforced concrete or lined with impermeable lining and are treated with anti-corrosion measures. They have sufficient volume to handle peak leachate volumes and are equipped with level gauges and booster pumps for subsequent collection and transfer.

[0032] Similarly, a landfill gas venting system is indispensable. During the anaerobic biodegradation of organic matter in waste, a large amount of landfill gas is generated, mainly composed of methane (CH4) and carbon dioxide (CO2), with methane being a potent greenhouse gas. To prevent uncontrolled gas escape that could cause environmental pollution, odor nuisance, or even explosion risks, a gas venting system must be constructed simultaneously during the landfill process. This typically includes vertical or horizontal gas collection wells, constructed of perforated pipes that penetrate the waste pile to collect and transport the gas to a main pipeline, which then converges at gas treatment facilities (such as flare towers, power generation units, or activated carbon adsorbers) for processing and utilization. Although this invention primarily focuses on leachate control, the composition of landfill gas is a key parameter for determining the degradation stage; therefore, gas collection and monitoring are an integral part of this ambitious concept.

[0033] After the landfill site is constructed, it begins receiving and filling municipal solid waste, following strict layered compaction procedures. Municipal solid waste is typically weighed and recorded before entering the landfill, and sometimes undergoes simple sorting to remove large items or hazardous materials. Once transported to the landfill, the waste is layered and compacted using bulldozers or compactors. Each layer is typically 0.5 to 2.0 meters thick; high compaction improves the stability of the waste pile, reduces settling, and effectively isolates air, promoting the formation of an anaerobic environment. The compacted waste layer is then covered with a covering material (such as clay, sand, or specialized foam) to reduce odor escape, prevent vector growth, and reduce the risk of fire. This process is repeated until the entire landfill is full.

[0034] After landfilling, the waste begins its biodegradation process within the storage area. Organic components such as food scraps, paper, and wood are gradually decomposed and transformed by the rich microbial community within the storage area. This process encompasses multiple biochemical stages, rapidly transitioning from initial aerobic decomposition (when a small amount of air remains in the waste) to facultative anaerobic decomposition, and then entering the crucial anaerobic decomposition stage. In anaerobic decomposition, hydrolytic microorganisms break down complex organic matter into simple organic acids, alcohols, and other small molecules. Acid-producing bacteria then further convert these into volatile fatty acids and hydrogen, causing a decrease in the pH of the leachate (acidification stage). Next, methanogenic bacteria become active, using these acid-producing products to further convert them into methane and carbon dioxide, raising the pH to neutral or slightly alkaline (methanogenesis stage). Finally, the waste enters a slow maturation stage, and the remaining recalcitrant organic matter gradually stabilizes.

[0035] Leachate generation is an inevitable and continuous phenomenon throughout the entire process of waste biodegradation. Its main sources include: the release of water from the waste itself under pressure; infiltration of external rainfall or surface water through the waste pile; and metabolic water produced during the degradation of organic matter. As this water flows downwards, it continuously dissolves and carries away soluble organic matter, inorganic salts, heavy metals, pathogens, and metabolic products from the waste, thus forming a high-concentration, complex leachate. The yield and quality of leachate dynamically vary depending on landfill time, waste composition, climatic conditions, and the stage of degradation. Initially, leachate production is typically high, with a high concentration of organic matter.

[0036] To efficiently collect leachate, this application incorporates a leachate drainage system (a gravel drainage layer and blind pipes) at the bottom of the storage area. Under gravity, the leachate flows along the sloping impermeable layer to the lowest point of the storage area, then passes through the drainage layer into the porous collection blind pipes. These blind pipes collect the leachate and channel it into a main collection pipe, ultimately leading to a leachate equalization tank located outside or within the storage area. The equalization tank is typically equipped with a level monitoring system to monitor the leachate level in real time and automatically activates a booster pump when the level reaches a preset high point, transferring the leachate to subsequent treatment facilities.

[0037] The above-described implementation methods can effectively ensure the safe isolation of the surrounding environment during the waste treatment process, providing a stable, reliable, and representative material basis for subsequent intelligent phased control based on leachate reinjection, and laying a solid foundation for improving waste degradation efficiency and shortening the stabilization cycle.

[0038] In the aforementioned intelligent solid waste environmental protection treatment method, step S2 involves acquiring the time series of biodegradation reaction state parameters within a predetermined time window at a predetermined sampling frequency. These biodegradation reaction state parameters include leachate pH, CH4 concentration, CO2 concentration, H2 concentration, and O2 concentration in the landfill gas. It should be understood that the biodegradation process of waste is a complex, dynamically evolving biochemical system whose internal state changes continuously over time. Therefore, in order to comprehensively and continuously capture and characterize the real-time dynamic changes of the internal environment of the waste pile, this application deploys high-precision online sensor arrays at key nodes in the leachate collection pond and landfill gas drainage network of the landfill area. These arrays are used to collect leachate pH (indicating acid-base balance), CH4 concentration (a marker of methanogenic bacteria activity), CO2 concentration (an indicator of total metabolic intensity), H2 concentration (a byproduct of the acidification stage), and O2 concentration (a criterion for aerobic / anaerobic states), thereby constructing a complete monitoring system for biodegradation reaction state parameters. In the specific implementation process, this application installs an industrial-grade pH electrode in the inlet of the leachate conditioning tank to continuously monitor the pH of the collected leachate; integrates an infrared gas analyzer (for detecting CH4 and CO2 concentrations) and an electrochemical sensor (for detecting H2 and O2 concentrations) in the gas collection main pipe of the landfill gas drainage network or a specific monitoring well, and performs automatic measurements according to a preset sampling frequency (e.g., once per hour) to obtain the pH value of the leachate and the components of the landfill gas within a predetermined time window (e.g., 30 days), forming a time series of biodegradation reaction state parameters containing multiple variables, thereby constructing a multi-dimensional, high-resolution degradation state evolution map, providing complete input for subsequent degradation stage identification.

[0039] In the aforementioned intelligent solid waste environmental protection treatment method, step S3 involves inputting the time series of the biodegradation reaction state parameters and the current reaction time into a waste biodegradation reaction stage identification model to obtain the current waste biodegradation reaction stage. More specifically, the current waste biodegradation reaction stage is an aerobic decomposition stage, an anaerobic decomposition stage, an acid-producing stage, a methanogenic stage, or a composting stage. It should be understood that, due to the dual influence of the historical state accumulation effect and the current reaction timescale on landfill biodegradation stage transitions, simple time threshold determination or single state parameter analysis can easily lead to misjudgments of stage transitions. Therefore, in order to accurately locate the current biodegradation stage (aerobic / anaerobic / acid-producing / methanogenic / composting), this application, based on the nonlinear temporal pattern modeling capability of deep learning for multidimensional time series data, constructs a waste biodegradation reaction stage identification model to comprehensively consider the temporal evolution law and timescale effect of historical waste biodegradation state data, thereby achieving accurate identification of the current waste biodegradation reaction stage.

[0040] Figure 2This is a flowchart of step S3 in the intelligent solid waste environmental protection treatment method according to an embodiment of this application. Figure 3 This is a schematic diagram of the data flow in step S3 of the intelligent solid waste environmental protection treatment method according to an embodiment of this application. Figure 2 and Figure 3 As shown, step S3 includes: S31, extracting the temporal change features of the time series of the biodegradation reaction state parameters to obtain the time series of the biodegradation reaction state temporal feature vector; S32, performing multi-level temporal context information transfer encoding on the time series of the biodegradation reaction state temporal feature vector to obtain the waste biodegradation reaction state feature temporal context transfer encoding vector; S33, inputting the current reaction time and the waste biodegradation reaction state feature temporal context transfer encoding vector into the trained classifier model to obtain the identification result of the current waste biodegradation reaction stage.

[0041] Specifically, step S31 involves extracting the temporal variation features of the time series of the biodegradation reaction state parameters to obtain a time series of the biodegradation reaction state temporal feature vector. In a specific example of this application, the time series of the biodegradation reaction state parameters is time-series modeled using a forward LSTM model to obtain a time series of the biodegradation reaction state temporal feature vector. It should be understood that waste biodegradation is a continuous biochemical process, and the current state (such as gas concentration and pH value) depends not only on the current instantaneous conditions but also on the evolutionary trend over a past period. For example, a continuous decrease in oxygen concentration indicates the imminent end of the aerobic stage, while a nascent increase in methane concentration marks the beginning of the methanogenesis stage. In other words, the temporal dependence pattern of the biodegradation reaction state parameters is key to determining the degradation stage. Therefore, in order to extract the dynamic evolution law of the biodegradation state from the time series of the biodegradation reaction state parameters, this application employs a Long Short-Term Memory (LSTM) network model for forward temporal modeling. Specifically, Long Short-Term Memory (LSTM) networks are a special type of Recurrent Neural Network (RNN) particularly suitable for handling long-term dependencies in time series. By introducing three control gates—input gate, forget gate, and output gate—into a standard RNN unit, they can effectively solve the gradient vanishing or gradient explosion problems that traditional RNNs often encounter when processing long-sequence data, thus achieving effective capture of long-distance dependencies in time series. In this application, multidimensional biodegradation reaction state parameters (leached leachate pH, CH4 concentration, CO2 concentration, H2 concentration, and O2 concentration in landfill gas) at various time points are arranged into a multidimensional state parameter input vector. The time series of this multidimensional state parameter input vector is then input into a feedforward LSTM model. The gating mechanism of the LSTM model is used to learn the long-term dependencies of each state parameter over time during the biodegradation process. By iteratively updating the hidden states at each time step, a time series of biodegradation reaction state time series feature vectors that integrate the current time step input and historical sequence information is obtained, providing a more accurate and comprehensive feature representation for subsequent identification of waste biodegradation reaction stages.

[0042] Specifically, in step S32, the time series of the biodegradation reaction state temporal feature vector is encoded using multi-level temporal context information transfer to obtain the waste biodegradation reaction state feature temporal context transfer encoding vector. It should be understood that, considering the characteristics of waste biodegradation stage transitions involving a mix of local abrupt changes and long-term gradual changes—for example, a sudden increase in CH4 concentration during the methanogenesis phase may be accompanied by a rapid transition from the acidogenesis phase to the methanogenesis phase, while a sustained high CH4 concentration for several consecutive days implies a stable and continuous methanogenesis phase—this means that while the LSTM model can capture the long-term trends of multidimensional state parameters over time when extracting the long-term dependencies of waste biodegradation states, it easily overlooks key local state changes. Therefore, in order to effectively integrate the local detailed features and global stage evolution patterns of the waste biodegradation process, this application proposes a multi-level temporal context information transfer encoding method. By performing local neighborhood context enhancement perception on the biodegradation reaction state temporal feature vectors at each time step in the time series of the biodegradation reaction state temporal feature vectors, the method focuses on the short-term event associations of the biodegradation process. Then, through global information transfer encoding, the method achieves deep integration of global temporal context information to obtain the waste biodegradation reaction state feature temporal context transfer encoding vector.

[0043] Figure 4 This is a flowchart of step S32 in the intelligent solid waste environmental protection treatment method according to an embodiment of this application. Figure 4 As shown, step S32 includes: S321, performing feature distribution sensing on each biodegradation reaction state time-series feature vector in the time series of the biodegradation reaction state time-series feature vector to determine the window size of the local context-enhanced sensing window for each biodegradation reaction state time-series feature vector; S322, based on all biodegradation reaction state time-series feature vectors in the local context-enhanced sensing window of each biodegradation reaction state time-series feature vector, performing local neighborhood context-enhanced sensing on each biodegradation reaction state time-series feature vector to obtain the time series of the local context-enhanced encoding vector of the waste biodegradation reaction state feature; S323, performing global temporal transfer encoding on the time series of the local context-enhanced encoding vector of the waste biodegradation reaction state feature to obtain the time series context-transfer encoding vector of the waste biodegradation reaction state feature.

[0044] In a specific example of this application, step S321 can be expressed by the formula:

[0045]

[0046] Where exp(·) represents the exponential function operation with base e, h i and hj Let |i| represent the i-th and j-th biodegradation reaction state time series feature vectors, respectively; ||·|| denotes the Euclidean norm calculation; τ represents the temperature coefficient; and s ij h i and h j The degree of correlation between features Let ∈ be the predefined neighborhood, γ represent the first smoothing term, γ represent the second smoothing term, and k represent the vector index. Indicates the preset neighborhood The temporal feature vector of the j-th biodegradation reaction state relative to h i The correlation weight coefficient, log2(·) represents the logarithmic function to the base 2, ε i Represents vector h i The nearest neighbor feature distribution entropy, w max Indicates the preset maximum window size, w i Represents vector h i The window size of the corresponding local context-enhanced perception window.

[0047] Here, the dynamic nature of the waste degradation process leads to spatiotemporal heterogeneity in the feature evolution rate. For example, the pH value can drop sharply by 2.0 units within 48 hours during the acidification stage, while the parameter change rate during the maturation stage is less than 0.1 / day. This makes it difficult for a fixed-size context window to adapt to the semantic association requirements of different stages / time points. Therefore, in order to dynamically capture the local environment most closely related to the temporal feature vectors of biodegradation reaction states at each time step, this application further performs feature distribution sensing on each biodegradation reaction state temporal feature vector. Based on its feature correlation degree and nearest neighbor feature distribution entropy relative to other biodegradation reaction state temporal feature vectors in a preset neighborhood, the window size of its corresponding local context enhancement sensing window is adaptively adjusted. This allows for accurate locking of the most informative local context environment under different biodegradation stages and parameter change rates.

[0048] In a specific example of this application, step S322 includes: inputting all biodegradation reaction state temporal feature vectors in the local context-enhanced perception window of the biodegradation reaction state temporal feature vector into a local neighborhood context-enhanced fusion network based on an attention mechanism to obtain the local context-enhanced encoding vector of the waste biodegradation reaction state features, expressed by the formula:

[0049]

[0050] Where softmax(·) represents the normalization exponential function, W q W k and W gThis represents the different weight parameter matrices in the local neighborhood context-enhanced fusion network, (·). T d represents the transpose of a vector, and d represents h. i Feature dimension, α ij h j The corresponding attention weight factor, Sigmoid(·) represents the sigmoid activation function. h i The corresponding local context-enhanced encoding vector of waste biodegradation reaction state features. Time series representing local context-enhanced encoded vectors of waste biodegradation reaction state characteristics. and They represent The local context-enhanced encoding vectors of the first, second, and i-th waste biodegradation reaction state features.

[0051] Here, considering that traditional sliding window average pooling context information fusion methods can obscure key mutation features (such as the pulse peak of H2 concentration during the transition from facultative oxalate to acid production), this application employs an attention-driven neighborhood context information aggregation method to achieve semantically enhanced feature fusion within the local context-enhanced perception window. By constructing a local neighborhood context-enhanced fusion network, saliency scoring and attention-weighted fusion are applied to all biodegradation reaction state time-series feature vectors within the local context-enhanced perception window corresponding to each biodegradation reaction state time-series feature vector. This highlights key features and suppresses non-key information, resulting in a time series of waste biodegradation reaction state feature local context-enhanced encoding vectors that incorporate local neighborhood context-enhanced information.

[0052] Specifically, considering the temperature coefficient τ and the first smoothing term ∈ and the second smoothing term γ introduced in the window size determination process of the local context-enhanced perception window, therefore, for W in the local neighborhood context-enhanced fusion network... g W q and W kPreferably, deformation feature quantification analysis should also be further introduced to improve the contextual semantic enhancement fusion effect under the predetermined local context enhancement perception window. Based on this, in a preferred example of this application, step S322 includes: first, optimizing the weight parameter matrix of the local neighborhood context enhancement fusion network based on the local neighborhood perception field to obtain an updated weight parameter matrix; then, updating the local neighborhood context enhancement fusion network based on the updated weight parameter matrix, and inputting all biodegradation reaction state time-series feature vectors in the local context enhancement perception window of the biodegradation reaction state time-series feature vector into the updated local neighborhood context enhancement fusion network for attention fusion to obtain the local context enhancement encoding vector of the waste biodegradation reaction state feature.

[0053] Specifically, firstly, for the deformation-related parameter set (τ,∈,γ) in the local context-enhanced sensing process of the biodegradation reaction state, linear interpolation is performed to make it the same length as the temporal feature vector of the biodegradation reaction state. Then, nonlinear curvature modeling is performed through one-dimensional convolution to generate the local context-enhanced sensing deformation feature vector V of the biodegradation reaction state. α And respectively the weight parameter matrix W g W q and W k As the mapping target, obtain the local context-enhanced perceptual curvature deformation correlation feature vector of the biodegradation reaction state:

[0054] V α1 =V α W g +V α

[0055] V α2 =V α W q +V α

[0056] V α3 =V α W k +V α

[0057] Among them, V α V represents the local context-enhanced perceptual deformation feature vector representing the state of a biodegradation reaction. α1 V α2 and V α3 These represent the weight parameter matrix W respectively. g W q and W k The corresponding biodegradation reaction state local context-enhanced perception curvature deformation associated feature vector.

[0058] Then, considering the direct coupling relationship with curvature deformation, the curvature-dominated weighted deformation can be further strengthened, i.e., ignoring the non-intrinsic higher-order nonlinear deformation components. Therefore, the aforementioned local context-enhanced perception curvature deformation-related feature vectors of the biodegradation reaction state are respectively coupled with the eigenvectors of the weight parameter matrix, i.e., V in the following equation. eg V eq and V ek To perform the correlation and correct the weight parameter matrix, it is represented as follows:

[0059] W' g =(V α1 T V eg )⊙W g

[0060] W' q =(V α2 T V eq )⊙W q

[0061] W' k =(V α3 T V ek )⊙W k

[0062] Among them, V eg V eq and V ek W g W q and W k The eigenvectors, ⊙ denotes the dot product operation, W' g W' q and W' k W g W q and W k The corresponding updated weight parameter matrix.

[0063] This allows the local context semantic enhancement fusion based on the weight parameter matrix to take into account the deformation factor introduced in the determination process of the local neighborhood sensing field, thereby improving the local context semantic enhancement fusion effect of all biodegradation reaction state time sequence feature vectors in the predetermined local context enhancement sensing window.

[0064] In a specific example of this application, step S323 includes: inputting the time series of the local context-enhanced encoding vector of the waste biodegradation reaction state features into a global transitive encoding network based on the Transformer architecture to obtain the time series context-enhanced encoding vector of the waste biodegradation reaction state features, expressed by the formula:

[0065]

[0066] Where Transformer(·) represents the Transformer architecture, v f The temporal context transmission encoding vector represents the state characteristics of waste biodegradation reaction.

[0067] It is understandable that the time series of the local context-enhanced encoding vector of the waste biodegradation reaction state features, even after local neighborhood enhancement, still suffers from the problem of missing long-range dependencies across stages. Therefore, in order to construct a deep feature representation that covers global temporal context evolution information, this application further introduces the Transformer architecture to perform global temporal context information transfer encoding on the time series of the local context-enhanced encoding vector of the waste biodegradation reaction state features. The Transformer architecture is known for its powerful global information capture capability. Based on a self-attention mechanism, it enables each time step feature to interact with other time step features in the sequence, thereby achieving deep integration of global temporal context information and effectively mining the global dependencies between time steps in the time series. In this application, the local context-enhanced encoding vector of the waste biodegradation reaction state features at each time step is used as the input of the Transformer architecture. The correlation score between each time step is calculated using a self-attention mechanism, and the features of each time step are weighted and summed based on this. This yields a waste biodegradation reaction state feature temporal context transfer encoding vector that integrates global temporal context information. This not only makes up for the shortcomings of the LSTM model in capturing local mutation features, but also achieves deep integration of global temporal context information, providing a more comprehensive and accurate feature representation for subsequent identification of waste biodegradation reaction stages.

[0068] Specifically, in step S33, the current reaction time and the temporal context transfer encoding vector of the waste biodegradation reaction state feature are input into the trained classifier model to obtain the identification result of the current waste biodegradation reaction stage. It should be understood that the reaction stage of waste biodegradation depends not only on the current state but also on the biological process constraints of the cumulative reaction time (e.g., the methanogenesis period typically occurs 90-150 days after landfill). Therefore, in order to effectively couple the temporal dynamic features of the waste biodegradation reaction state with the inherent temporal patterns of biodegradation, thereby avoiding confusion between similar local dynamic patterns appearing in different degradation stages, this application, based on the principle of multimodal information fusion decision-making, integrates the current reaction time of waste biodegradation into the temporal context transfer encoding vector of the waste biodegradation reaction state feature, and inputs it into the trained classifier model for joint decision-making. This allows the model to comprehensively consider the time accumulation effect and current state features during the waste biodegradation process, achieving accurate segmentation of the waste biodegradation reaction stage. Specifically, firstly, the current reaction time is standardized to the [0,1] interval and then concatenated with the temporal context transfer encoding vector of the waste biodegradation reaction state feature to form a fused feature vector containing time dimension information, which is then input into a pre-trained classifier model. This classifier model is based on a fully connected neural network architecture and uses a multilayer perceptron to perform nonlinear mapping on the input features. Finally, the Softmax layer outputs the probability distribution of the current waste biodegradation in five stages: aerobic, facultative, acidification, methanogenesis, and decomposition. The stage corresponding to the highest probability value is selected as the identification result of the current waste biodegradation reaction stage, thereby achieving real-time monitoring and precise management of the waste biodegradation process.

[0069] In the aforementioned intelligent solid waste environmental protection treatment method, step S4 determines the target pH value of the reinjected leachate based on the current stage of waste biodegradation reaction. It should be understood that the optimal pH range for the dominant microbial communities at different degradation stages differs significantly (e.g., acid-producing bacteria are acid-tolerant while methanogens require near-neutral pH). Therefore, in order to create the optimal metabolic environment for microorganisms at each stage of biodegradation reaction, this application establishes a mapping rule between biodegradation reaction stages and pH values ​​based on the optimal relationship between microbial enzyme activity and pH value, to determine the optimal pH value matching the current stage of waste biodegradation reaction. In a specific example of this application, if the current waste biodegradation reaction stage is the aerobic decomposition stage, the target pH value of the recharge leachate is set to 7.0; if the current waste biodegradation reaction stage is the facultative anaerobic decomposition stage, the target pH value of the recharge leachate is set to 7.5; if the current waste biodegradation reaction stage is the acidification stage, the target pH value of the recharge leachate is set to 8.0; if the current waste biodegradation reaction stage is the methanogenesis stage, the target pH value of the recharge leachate is set to 7.0; if the current waste biodegradation reaction stage is the composting stage, no recharge leachate is required, and the landfill area is maintained in its natural state. Based on this, after determining the current waste biodegradation reaction stage, by querying a preset table of correspondence between biodegradation reaction stages and pH values, the optimal pH value matching the current waste biodegradation reaction stage can be quickly and accurately obtained, thereby achieving precise matching between pH regulation and biological phase change, maximizing the activity of key functional microbial communities, and optimizing the microbial metabolic environment.

[0070] In the aforementioned intelligent solid waste environmental protection treatment method, step S5 involves adjusting the pH of the leachate based on the deviation between the target pH value of the recharge leachate and the current pH value of the leachate to obtain recharge leachate, and then recharging the recharge leachate back into the landfill. It should be understood that leachate recharge requires real-time matching with dynamically optimized pH conditions. Therefore, to translate the determined optimal control strategy (i.e., the target pH value of the recharge leachate) into a physical control action, this application uses a PID control algorithm based on the principle of acid-base neutralization to adjust the pH of the leachate in the leachate conditioning tank in real time. Specifically, after the collected raw leachate is pumped into the leachate conditioning tank equipped with a stirring device, a pH sensor in the tank measures the current pH value of the leachate in real time. The PID (Proportional-Integral-Derivative) controller continuously compares the deviation between the current leachate pH value and the target pH value of the reinjected leachate. Based on this deviation, it calculates the dosage of an acid-base regulator (usually lime water, sodium hydroxide, or other alkaline agents) and instructs the metering pump to pump the acid-base regulator into the leachate conditioning tank according to the calculated dosage. The mixture is then thoroughly stirred until the leachate pH value reaches the preset target pH value for the reinjected leachate. After conditioning, the resulting reinjected leachate is evenly or selectively sprayed or injected into the top or interior of the landfill through a dedicated pump and pipeline system. This direct intervention in the biochemical environment of the landfill creates optimal pH conditions for the dominant microbial communities at specific stages, ultimately accelerating waste degradation, shortening the stabilization period, and improving overall treatment efficiency.

[0071] In summary, the intelligent solid waste environmental treatment method according to the embodiments of this application is explained. It involves landfilling municipal solid waste and collecting leachate generated during its biodegradation process. Simultaneously, it monitors key state parameters such as pH value and landfill gas composition of the leachate in real time during the biodegradation process. By introducing a deep learning algorithm, it analyzes the temporal evolution patterns of multidimensional key state parameters during the waste biodegradation process. Furthermore, it adaptively identifies the current stage of waste biodegradation reaction based on the current reaction time. Subsequently, based on the identification results of the reaction stage, it determines a suitable target pH value for the reinjected leachate and adjusts the pH of the collected leachate accordingly. The adjusted leachate is then reinjected into the landfill area to dynamically regulate its pH environment. This method, through intelligent sensing and phased control of the waste degradation process, can provide an optimal biochemical environment for microbial communities at different stages, thereby improving waste degradation efficiency and gas production rate.

[0072] Furthermore, this application also provides an intelligent solid waste environmental protection treatment system.

[0073] Figure 5 This is a block diagram of an intelligent solid waste environmental protection treatment system according to an embodiment of this application. Figure 5 As shown, the intelligent solid waste environmental protection treatment system 100 according to an embodiment of this application includes: a leachate collection module 110, used to landfill municipal solid waste in a landfill area for biodegradation and collect the generated leachate; a biodegradation reaction status monitoring module 120, used to acquire the time series of biodegradation reaction status parameters within a predetermined time window at a predetermined sampling frequency, the biodegradation reaction status parameters including leachate pH value, CH4 concentration, CO2 concentration, H2 concentration and O2 concentration in landfill gas; a biodegradation reaction stage identification module 130, used to input the time series of the biodegradation reaction status parameters and the current reaction time into a waste biodegradation reaction stage identification model to obtain the current waste biodegradation reaction stage; a target pH value determination module 140, used to determine the target pH value of the recharge leachate based on the current waste biodegradation reaction stage; and an acidity / alkalinity adjustment module 150, used to adjust the acidity / alkalinity of the leachate based on the deviation between the target pH value of the recharge leachate and the current pH value of the leachate to obtain the recharge leachate, and recharge the recharge leachate back to the landfill area.

[0074] Here, those skilled in the art will understand that the specific operation of each module in the aforementioned intelligent solid waste environmental protection treatment system has been described in the above-mentioned... Figures 1 to 4 The intelligent solid waste environmental protection treatment method is described in detail in the description, and therefore, its repeated description will be omitted.

[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An intelligent environmentally friendly method for treating solid waste, characterized in that, include: Municipal waste is landfilled in landfill areas for biodegradation, and the resulting leachate is collected; The time series of biodegradation reaction state parameters within a predetermined time window are obtained at a predetermined sampling frequency. The biodegradation reaction state parameters include leachate pH value, CH4 concentration, CO2 concentration, H2 concentration and O2 concentration in landfill gas. The time series of the biodegradation reaction state parameters and the current reaction time are input into the waste biodegradation reaction stage identification model to obtain the current waste biodegradation reaction stage. Based on the current stage of waste biodegradation reaction, the target pH value of the reinjection leachate is determined; The pH of the leachate is adjusted based on the deviation between the target pH value and the current pH value of the recharge leachate to obtain the recharge leachate, and the recharge leachate is then recharged into the landfill.

2. The intelligent solid waste environmental protection treatment method according to claim 1, characterized in that, The current waste biodegradation reaction stage is described as aerobic decomposition, facultative anaerobic decomposition, acid production, methanogenesis, or composting.

3. The intelligent solid waste environmental protection treatment method according to claim 2, characterized in that, Based on the current stage of waste biodegradation, the target pH value of the reinjected leachate is determined, including: If the current stage of waste biodegradation is the aerobic decomposition stage, the target pH value of the reinjected leachate is set to 7.0; If the current stage of waste biodegradation is the facultative anaerobic decomposition stage, the target pH value of the reinjected leachate is set to 7.5; If the current stage of waste biodegradation is the acid production stage, the target pH value of the reinjected leachate is set to 8.0; If the current stage of waste biodegradation is the methanogenesis stage, the target pH value of the reinjected leachate is set to 7.0; If the current stage of waste biodegradation is the maturation stage, there is no need to refill leachate, thus maintaining the natural state of the landfill area.

4. The intelligent solid waste environmental protection treatment method according to claim 2, characterized in that, The time series of the biodegradation reaction state parameters and the current reaction time are input into the waste biodegradation reaction stage identification model to obtain the current waste biodegradation reaction stage, including: The temporal variation features of the time series of the biodegradation reaction state parameters are extracted to obtain the time series of the biodegradation reaction state temporal feature vector; Multi-level temporal context information transfer encoding is performed on the time series of the biodegradation reaction state temporal feature vector to obtain the waste biodegradation reaction state feature temporal context transfer encoding vector; The current reaction time and the temporal context of the waste biodegradation reaction state feature are passed to the encoded vector and input into the trained classifier model to obtain the identification result of the current waste biodegradation reaction stage.

5. The intelligent solid waste environmental protection treatment method according to claim 4, characterized in that, Extracting the temporal variation features of the time series of the biodegradation reaction state parameters to obtain the time series of biodegradation reaction state temporal feature vectors includes: The time series of the biodegradation reaction state parameters are modeled using a forward LSTM model to obtain the time series of the biodegradation reaction state time series feature vector.

6. The intelligent solid waste environmental protection treatment method according to claim 5, characterized in that, Multi-level temporal context information transfer encoding is performed on the time series of the biodegradation reaction state temporal feature vector to obtain the waste biodegradation reaction state feature temporal context transfer encoded vector, including: The feature distribution perception is performed on each biodegradation reaction state time-series feature vector in the time series of the biodegradation reaction state time-series feature vector to determine the window size of the local context-enhanced perception window for each biodegradation reaction state time-series feature vector; Based on all biodegradation reaction state time-series feature vectors in the local context-enhanced perception window of each biodegradation reaction state time-series feature vector, local neighborhood context-enhanced perception is performed on each biodegradation reaction state time-series feature vector to obtain the time series of the local context-enhanced encoding vector of waste biodegradation reaction state features. Global temporal transfer encoding is performed on the time series of the local context-enhanced encoding vector of the waste biodegradation reaction state features to obtain the temporal context transfer encoding vector of the waste biodegradation reaction state features.

7. The intelligent solid waste environmental protection treatment method according to claim 6, characterized in that, Based on all biodegradation reaction state time-series feature vectors within the local context-enhanced sensing window of each biodegradation reaction state time-series feature vector, local neighborhood context-enhanced sensing is performed on each biodegradation reaction state time-series feature vector to obtain the time series of the local context-enhanced encoded vector of the waste biodegradation reaction state features, including: All biodegradation reaction state temporal feature vectors in the local context-enhanced perception window are input into a local neighborhood context-enhanced fusion network based on an attention mechanism to obtain the local context-enhanced encoding vector of the waste biodegradation reaction state features.

8. The intelligent solid waste environmental protection treatment method according to claim 7, characterized in that, The local context-enhanced feature vectors of the biodegradation reaction state temporal feature vectors are input into a local neighborhood context-enhanced fusion network based on an attention mechanism to obtain the local context-enhanced encoding vector of the waste biodegradation reaction state features, including: The weight parameter matrix of the local neighborhood context enhancement fusion network is optimized by weight parameter adaptation based on the local neighborhood perception field to obtain an updated weight parameter matrix; The local neighborhood context enhancement fusion network is updated based on the updated weight parameter matrix, and all biodegradation reaction state time-series feature vectors in the local context enhancement perception window of the biodegradation reaction state time-series feature vector are input into the updated local neighborhood context enhancement fusion network for attention fusion to obtain the local context enhancement encoding vector of the waste biodegradation reaction state feature.

9. The intelligent solid waste environmental protection treatment method according to claim 6, characterized in that, Global temporal transfer encoding is performed on the time series of the local context-enhanced encoding vector of the waste biodegradation reaction state features to obtain the temporal context-transfer encoding vector of the waste biodegradation reaction state features, including: The time series of the local context-enhanced encoding vector of the waste biodegradation reaction state features is input into a global transitive encoding network based on the Transformer architecture to obtain the temporal context-enhanced encoding vector of the waste biodegradation reaction state features.

10. An intelligent solid waste environmental protection treatment system, characterized in that, include: The leachate collection module is used to collect leachate generated from municipal solid waste that has been biodegraded in landfill areas. A biodegradation reaction status monitoring module is used to acquire the time series of biodegradation reaction status parameters within a predetermined time window at a predetermined sampling frequency. The biodegradation reaction status parameters include leachate pH value, CH4 concentration, CO2 concentration, H2 concentration and O2 concentration in landfill gas. The biodegradation reaction stage identification module is used to input the time series of the biodegradation reaction state parameters and the current reaction time into the waste biodegradation reaction stage identification model to obtain the current waste biodegradation reaction stage. The target pH value determination module is used to determine the target pH value of the reinjection leachate based on the current stage of waste biodegradation reaction. A pH adjustment module is used to adjust the pH of the leachate based on the deviation between the target pH value of the recharge leachate and the current pH value of the leachate to obtain recharge leachate, and to recharge the recharge leachate back into the landfill area.

Images