Method and system for determining fly ash disposal path

Abstract

This application relates to the planning field and provides a method and system for determining fly ash disposal paths. It obtains the attribute parameters of the material to be allocated, the process requirement parameters of the target node, the rated capacity, the currently allocated load, and the unavailable capacity. A flow network is constructed, and the fit degree is calculated based on the attribute and process requirement parameters as a feasibility constraint to eliminate infeasible paths. The dynamic available capacity is determined based on the rated capacity, the currently allocated load, and the unavailable capacity as the upper limit of the allocation quantity constraint. Based on the feasibility constraint and the allocation quantity constraint, the flow network is optimized using a multi-objective method to determine the optimal allocation path. This application simultaneously injects fit degree constraints and dynamic available capacity constraints into the optimization solution, correcting the solution space boundary and achieving dual feasibility of the allocation path in terms of process matching and capacity carrying capacity. This significantly improves the engineering executability and stability of the optimization results.

Description

Technical Field

[0001] This application relates to the field of planning, and more specifically, to a method and system for determining the disposal path of fly ash. Background Technology

[0002] Currently, in the fields of resource allocation and route planning such as hazardous waste disposal and logistics scheduling, the route selection of materials to be allocated from the source to the target node usually relies on static parameters and experience-based decisions.

[0003] Existing solutions often involve simple point-to-point matching between the source and target nodes, and perform single-objective or simply weighted path optimization based on static design processing capabilities or historical average capabilities as constraints.

[0004] However, in actual engineering operations, there is a specific compatibility relationship between the attribute parameters (such as composition, physical properties, etc.) of different materials to be allocated and the process requirements parameters of the target node, and the available processing capacity of the target node is in dynamic change due to load changes, maintenance shutdowns or policy restrictions.

[0005] Existing solutions lack quantitative constraints on the compatibility of material properties with process requirements, which can easily lead to materials entering incompatible process paths, reducing processing efficiency or generating additional costs. At the same time, the lack of a mechanism for calculating the dynamic availability of target nodes can easily lead to problems such as nodes being overloaded or paths being infeasible in actual implementation of the optimized allocation results.

[0006] Therefore, there is an urgent need for a path optimization scheme that can simultaneously consider the feasibility of material-process adaptation and the feasibility of node dynamic capacity carrying capacity during the path optimization process. Summary of the Invention

[0007] The purpose of this application is to provide a method and system for determining fly ash disposal paths. In multi-node allocation optimization, material-process adaptation constraints and node dynamic availability constraints are injected simultaneously to correct the solution space boundary of the optimization model, realize the dual feasibility of allocation paths in terms of process matching and capacity carrying capacity, and improve the engineering executability and stability of optimization results.

[0008] Firstly, this application provides a method for determining the disposal path of fly ash, including: Obtain the attribute parameters of the materials to be allocated, the process requirements parameters of the target node, the rated capacity, the currently allocated load, and the unavailable capacity; Construct the flow network of the materials to be allocated to the target nodes; Based on the attribute parameters and the process requirement parameters, the fit between the material to be allocated and the target node is calculated, and the fit is used as a feasibility constraint to eliminate infeasible paths. Based on the rated capacity, the currently allocated load, and the unavailable capacity, the dynamic availability capacity of the target node is determined, and the dynamic availability capacity is used as the upper limit of the allocation constraint. Based on the feasibility constraints and the allocation constraints, the flow network is optimized using a multi-objective approach to determine the optimal allocation path.

[0009] In an optional implementation, the fit is used as a feasibility constraint to eliminate infeasible paths, including: Hard constraint: Boolean judgment based on qualification or category; if not satisfied, the path is disabled. Soft constraint: The path is disabled if the degree of matching between the attribute parameters and the process requirement parameters is less than the preset threshold.

[0010] In an optional implementation, the calculation of the quantized matching degree of the soft constraint includes: Calculate the degree of individual matching between the material to be allocated and the target node on each indicator; The overall matching degree is obtained by weighted summation of the matching degrees of each individual item; The degree of matching of each indicator is determined based on the correspondence between the attribute parameters and the process requirement parameters on that indicator, and the weighting coefficient reflects the importance of each indicator in the comprehensive evaluation.

[0011] In an optional implementation, determining the dynamic availability of the target node based on the rated capacity, the currently allocated load, and the unavailable capacity includes: Dynamic availability equals rated capacity minus currently allocated load minus unavailable capacity; The unavailable capacity includes capacity occupied due to maintenance, production restrictions, or policy limitations.

[0012] In an optional implementation, using the dynamically available capacity as an upper limit for allocation constraints includes: The total amount of materials allocated to the target node shall not exceed its current dynamic available capacity; The dynamic availability capability is updated dynamically over time.

[0013] In an optional implementation, the optimization objective of the multi-objective optimization solution includes: Minimize total cost; Minimize environmental impact.

[0014] In an optional implementation, determining the optimal allocation path includes: Obtain the Pareto optimal solution set obtained from multi-objective optimization; Calculate the distance between each solution and the ideal solution and the negative ideal solution; The relative proximity of each solution is calculated based on the distance. The Pareto optimal solution set is sorted according to relative proximity, and the optimal allocation path is selected.

[0015] Secondly, this application provides a device for determining the disposal path of fly ash, comprising: The acquisition module is used to acquire the attribute parameters of the materials to be allocated, the process requirement parameters of the target node, the rated capacity, the currently allocated load, and the unavailable capacity. A network construction module is used to construct the flow network of the materials to be allocated to the target nodes; The adaptability calculation module is used to calculate the adaptability between the material to be allocated and the target node based on the attribute parameters and the process requirement parameters, and to use the adaptability as a feasibility constraint to eliminate infeasible paths. The dynamic capacity determination module is used to determine the dynamic availability capacity of the target node based on the rated capacity, the currently allocated load, and the unavailable capacity, and to use the dynamic availability capacity as the upper limit of the allocation constraint. The optimization solution module is used to perform multi-objective optimization on the flow network based on the feasibility constraints and the allocation constraints to determine the optimal allocation path.

[0016] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for determining the fly ash disposal path as described in any of the foregoing embodiments.

[0017] Fourthly, this application provides a system for determining the disposal path of fly ash, including a processor configured to perform the method for determining the disposal path of fly ash as described in any of the foregoing embodiments.

[0018] Beneficial effects: By using the compatibility between material properties and process requirements as a feasibility constraint, infeasible processes are eliminated before optimization, thus avoiding efficiency reduction or additional costs caused by materials entering unsuitable processes and improving the process matching feasibility of the optimization results.

[0019] By calculating the dynamic availability of the target node based on rated capacity, currently allocated load, and unavailable capacity, and using this as the upper limit of allocation constraints, the limitations of traditional static capacity constraints are overcome, node overload is prevented, and the feasibility of capacity carrying capacity of the optimization results is improved.

[0020] By simultaneously injecting fitness constraints and dynamic availability constraints into the multi-objective optimization model, the solution space boundary of the optimization model is corrected, making the final determined optimal allocation path both feasible in terms of process matching and capacity carrying capacity, which significantly improves the executability and stability of the optimization results in the actual dynamic operation environment of engineering.

[0021] By dividing the fit constraint into hard and soft constraints, the system balances rigid compliance requirements with flexible quantitative evaluation, thereby improving the engineering flexibility of the system. By sorting the optimal path based on the relative proximity of the ideal solution distance, the system improves the comprehensive decision-making ability under multi-objective conflict conditions. Attached Figure Description

[0022] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 A schematic flowchart illustrating a method for determining the disposal path of fly ash, provided in an embodiment of this application; Figure 2 A schematic diagram of a device for determining the disposal path of fly ash provided in an embodiment of this application; Figure 3 This is a schematic diagram of a system structure for determining the disposal path of fly ash, provided in an embodiment of this application. Detailed Implementation

[0024] The technical solutions in the embodiments of this application will now be described with reference to the accompanying drawings.

[0025] This embodiment provides a method for determining fly ash disposal paths. This method aims to correct the solution space boundary of the optimization model by simultaneously injecting material and process compatibility constraints and node dynamic capacity allocation constraints into the multi-node allocation optimization model. This ensures that the final output optimal allocation path is feasible in both process matching and capacity carrying capacity, significantly improving the executability and stability of the optimization results in actual engineering dynamic operating environments. Figure 1 As shown, the method mainly includes the following steps: Step S110: Obtain the attribute parameters of the material to be allocated, the process requirement parameters of the target node, the rated capacity, the currently allocated load, and the unavailable capacity.

[0026] Specifically, materials to be allocated refer to various material flows that need to be allocated from the source of generation to the target node for processing, treatment or disposal, and their attribute parameters reflect the physical, chemical or category characteristics of the material itself.

[0027] The target node refers to various facilities or sites that receive and process the aforementioned materials. Its process requirement parameters describe the specific access requirements or processing preferences of the node for the incoming materials.

[0028] Rated capacity is the maximum theoretical processing capacity specified by the target node during the design and planning phase. The currently allocated load is the processing capacity that the node has committed to or actually occupied at the current moment. Unavailable capacity is the idle capacity that cannot be put into use due to various temporary or periodic factors.

[0029] It should be understood that although this embodiment uses fly ash and disposal facilities in a hazardous waste disposal scenario as a typical example for illustration, in other embodiments, the materials to be allocated may also be parcels in logistics scheduling, raw materials in production and manufacturing, or power loads in the energy network; correspondingly, the target node may also be a logistics transit warehouse, a production line station, or a substation, etc. As long as there are requirements for the compatibility between material properties and node processes, as well as dynamic fluctuation characteristics of node capabilities, this method can be applied.

[0030] Step S120: Construct the flow network of materials to be allocated to target nodes.

[0031] Specifically, the flow network serves as the environmental foundation and topological carrier for the subsequent multi-objective optimization model. This step abstracts each source and target node as a set of nodes in the network, and the feasible transportation paths between sources and targets as a set of edges, assigning attribute parameters such as distance, unit transportation cost, unit transportation carbon emissions, and transportation risk coefficient to the edges. Through this network-based modeling approach, the traditional simple point-to-point matching relationship is extended into a globally computable topological structure, enabling path selection to be optimized collaboratively in complex multi-source and multi-sink relationships.

[0032] This application does not limit the specific algorithm for constructing the flow network. Those skilled in the art can generate the required edge set by using the shortest path algorithm or spatial network analysis method based on actual traffic network data or spatial geographic information.

[0033] Step S130: Based on attribute parameters and process requirement parameters, calculate the fit between the material to be allocated and the target node, and use the fit as a feasibility constraint to eliminate infeasible paths.

[0034] Specifically, fit is a quantitative evaluation indicator of whether a material can enter a specific target node. Injecting fit as a feasibility constraint into the optimization model is the first line of defense in correcting the solution space boundary. If fit feasibility constraints are not introduced before multi-objective optimization, and free optimization is performed solely based on objective functions such as cost or distance, the boundary of the solution space will be in a virtually wide state. That is, the model might assign high-chlorine-content materials to process nodes only suitable for low-chlorine inputs, or forcibly match materials without specific qualification categories to facilities with strict access restrictions. Such optimization results, lacking process fit control, may theoretically be extreme values ​​of the objective function, but in actual engineering execution, they can lead to a sharp drop in processing efficiency, additional processing costs, and even safety accidents due to process incompatibility, ultimately rendering the project infeasible. Therefore, by calculating fit and setting a threshold, those absolutely infeasible path branches are rigidly eliminated from the solution space before the optimization process begins. This ensures that the subsequent optimization process only takes place within the truly feasible subspace, fundamentally guaranteeing the process matching feasibility of the optimization results.

[0035] Step S140: Based on the rated capacity, the currently allocated load, and the unavailable capacity, determine the dynamic availability capacity of the target node, and use the dynamic availability capacity as the upper limit of the allocation constraint.

[0036] Specifically, the formula for calculating dynamic availability capacity is: dynamic availability capacity equals rated capacity minus currently allocated load minus unavailable capacity.

[0037] Injecting dynamic available capacity as the upper limit of allocation constraints into the optimization model is the second line of defense for correcting the solution space boundary. If only static rated capacity is used as the allocation constraint, ignoring the unavailable capacity caused by maintenance shutdowns, temporary production restrictions, or policy limitations during facility operation, the solution space boundary will also be artificially wide or even false. The model may over-allocate materials to a target node that is under maintenance based on the rated upper limit, causing severe overload congestion or direct rejection at the actual receiving point, making the optimized allocation scheme impossible to implement.

[0038] By introducing unavailable capacity as a core variable, the available carrying capacity of the target node is modified from a static theoretical value to a dynamic value that reflects the current actual operating status. This dynamic value is then used as the upper limit of the allocation constraint, effectively preventing the problem of node overload allocation and ensuring the feasibility of the capacity carrying capacity of the optimization results.

[0039] Step S150: Based on feasibility constraints and allocation constraints, perform multi-objective optimization on the flow network to determine the optimal allocation path.

[0040] Specifically, after the fitness feasibility constraint elimination in step S130 and the dynamic available capacity allocation constraint limitation in step S140, the solution space boundary flowing to the network has been corrected to an effective boundary that truly reflects the dual feasibility of process matching and capacity carrying capacity. Within this corrected solution space, a multi-objective optimization process is performed. The optimization objectives typically encompass multiple conflicting dimensions such as minimizing total cost and minimizing environmental impact. This application does not limit the specific multi-objective optimization algorithm used. Those skilled in the art can choose non-dominated sorting genetic algorithms, multi-objective particle swarm optimization algorithms, decomposition-based multi-objective evolutionary algorithms, or other suitable multi-objective optimization algorithms to solve the problem, based on the problem size and real-time requirements, to obtain a Pareto optimal solution set that satisfies the dual constraints, and determine the final optimal allocation path from it.

[0041] Through the synchronous injection of the above dual constraints, this embodiment achieves dual feasibility of allocation path in terms of process matching and capacity carrying capacity, significantly improving the executability and stability of optimization results in the actual dynamic operation environment of engineering.

[0042] In some embodiments, the specific decision-making logic for calculating the fitness degree and using the fitness degree as a feasibility constraint to eliminate infeasible paths in step S130 is further elaborated. The fitness degree as a feasibility constraint to eliminate infeasible paths includes two levels: hard constraints and soft constraints.

[0043] Specifically, hard constraints are based on Boolean judgments of qualifications or categories; if the conditions are not met, the path is disabled.

[0044] The role of hard constraints is to rigidly intercept path branches that are absolutely mismatched in terms of compliance or basic physical category, ensuring that the optimization model does not allocate materials to illegal or infeasible nodes.

[0045] Boolean-type decisions mean that their output is either logically true or logically false, with no intermediate states. For example, in a hazardous waste disposal scenario, if a target node only has the qualification permit to process waste containing heavy metals, but the category label of the material to be allocated shows ordinary solid waste without heavy metals, or the target node is explicitly outside the administrative region where the material is generated, then the hard constraint decision result is false. In this case, the decision variable for that path is directly forced to zero, i.e. =0, cutting off the possibility of distributing the material to this node, making this path absolutely disabled in subsequent multi-objective optimization solutions.

[0046] It should be understood that although this embodiment uses qualification licenses and category labels as typical scenarios of hard constraints, in other embodiments, hard constraints can also be any basic access conditions with an either-or nature, such as physical compatibility determination of device interfaces, height and weight limits of transportation channels, etc.

[0047] Furthermore, soft constraints are calculated based on the quantitative matching degree between attribute parameters and process requirement parameters; if the degree is lower than a preset threshold, the path is disabled.

[0048] The role of soft constraints is to provide a flexible evaluation of the multi-dimensional, refined comparison between material attributes and process requirements, blocking paths that meet basic access criteria but have poor actual matching. The quantitative matching degree calculation of soft constraints includes calculating the individual matching degree of the material to be assigned and the target node on each indicator, and then weighting and summing the individual matching degrees to obtain the comprehensive matching degree.

[0049] The degree of matching for each indicator is determined based on the correspondence between the attribute parameters and the process requirement parameters for that indicator, and the weighting coefficient reflects the importance of each indicator in the comprehensive evaluation.

[0050] The formula for calculating the overall matching degree can be expressed as: ;in This represents the overall matching degree between the material i to be allocated and the target node j. This represents the weight coefficient of the k-th adaptation metric. This represents the single-matching function for the k-th indicator, used to describe the attribute parameter value. With process requirements The degree of matching between them.

[0051] When the calculated overall matching degree When the value is below the preset adaptation threshold σ, this path is also disabled. =0.

[0052] Regarding the implementation method of the one-to-one matching function φ, this application does not limit it to a single method; it can also be based on the calculation of the deviation ratio, for example... This means measuring the degree of matching by calculating the ratio of the absolute deviation between the actual attribute value and the required value to the required value; the smaller the deviation, the higher the degree of matching. Alternatively, an interval mapping function can be used, for example, when the attribute value... Falling within the optimal acceptable range set by the process requirements [ , When the value of φ is 1, when the value deviates from the range but is within the tolerable range, the value of φ decreases linearly. When the value exceeds the tolerable range, the value of φ is 0. In addition, a fuzzy membership function can be used to transform the matching relationship between attribute parameters and process requirement parameters into the membership calculation of fuzzy sets, so as to handle indicators with fuzzy boundaries more smoothly.

[0053] Because multiplication logic is overly sensitive to extreme weaknesses in a single indicator, even a slight deviation in one indicator can cause the overall matching degree to plummet to zero, easily leading to the over-elimination of acceptable paths; while minimum-value logic completely ignores the advantages and compensation of other dimensions. Weighted summation can effectively balance the conflict and complementarity between multi-dimensional indicators. It allows a slight deficiency in a non-core indicator to be compensated for by the advantages of other core indicators, while ensuring the decisive role of core indicators in the comprehensive evaluation through weighting coefficients. Moreover, it has high computational efficiency and adjustable parameters, making it very suitable for real-time optimization decision-making in engineering systems.

[0054] Furthermore, in actual engineering operations, due to limitations in detection methods or data transmission delays, certain attribute parameters of the materials to be allocated may be missing. To ensure the engineering flexibility and stability of the adaptation calculation mechanism, this embodiment also provides a mechanism for handling missing indicators. When some attribute indicators cannot be obtained in real time, the first step is to attempt to use historical detection data from the same source to impute the missing indicators with the mean or estimate the interval. Secondly, when similar samples from the same process, similar composition, or similar regions exist, similar sample substitution methods can be used to supplement them. Thirdly, a multiple regression model or machine learning prediction model can be established based on known indicators to estimate the missing indicators. For indicators that are still difficult to obtain, reasonable interval values ​​can be set based on the material's category label and expert experience rules. Through the above multi-level processing mechanism, the missing indicators are supplemented and expressed while ensuring the stable operation of the adaptation model, avoiding the problem of excessive elimination due to a single missing data point leading to the entire path being misjudged as infeasible.

[0055] This embodiment divides the fit constraint into hard constraints based on Boolean judgment and soft constraints based on quantitative calculation. This ensures the absolute satisfaction of rigid compliance requirements such as qualification categories, while providing a flexible evaluation space for the quantitative matching of attribute parameters and process requirements, thus improving the engineering feasibility and flexibility of fit judgment. At the same time, through the weighted summation of soft constraint quantitative logic and the flexible supplementation mechanism for missing indicators, a multi-dimensional and refined comparison between material attributes and process requirements is achieved, making the judgment logic of soft constraints more accurate and adjustable, and further improving the accuracy and robustness of path feasibility screening.

[0056] In some embodiments, the specific calculation and update mechanism for determining the dynamic availability of the target node and using the dynamic availability as the upper limit of the allocation constraint in step S140 is further elaborated.

[0057] Specifically, based on the rated capacity, the currently allocated load, and the unavailable capacity, the dynamic availability capacity of the target node is determined, including: the dynamic availability capacity equals the rated capacity minus the currently allocated load minus the unavailable capacity.

[0058] This accounting logic can be expressed by the following formula: ,in This represents the dynamic available capability of target node j at time t. This represents the rated capacity of target node j. This represents the current allocated load of target node j at time t. This represents the unavailable capacity of target node j at time t.

[0059] In this formula, unavailable capacity It is the source of core variables that distinguishes it from traditional static capability models.

[0060] Unavailable capacity includes capacity occupied due to maintenance, production restrictions, or policy limitations.

[0061] For example, in the hazardous waste disposal scenario, maintenance occupation refers to the shutdown of disposal facilities due to regular maintenance or sudden equipment failure, during which the facilities, although physically present, completely lose their processing capacity; production restriction occupation refers to the facility actively reducing its operating load due to process adjustments, raw material shortages, or fluctuations in the downstream product market, resulting in some of its designed capacity being idle; policy restriction occupation refers to the facility being forced to suspend receiving materials of specific categories or sources due to administrative factors such as environmental protection inspections and rectifications, local temporary control orders, or freezes on cross-regional transfer approvals.

[0062] It should be understood that although this embodiment lists three typical scenarios—maintenance, production restrictions, and policy limitations—in other embodiments, the sources of unavailable capacity may also include any temporary or periodic factors that prevent the rated capacity from being actually put into use, such as shutdowns due to extreme weather, shutdowns due to material shortages caused by upstream supply disruptions, or forced shutdowns due to safety production accidents.

[0063] If the optimization model uses only static rated capacity as the constraint boundary, the boundary of the solution space will be in a state of spurious abundance. The model is highly likely to over-allocate materials to a target node that is under maintenance shutdown or production restricted by policy, based on the upper limit of this spurious abundance. This allocation scheme may be optimal in terms of mathematical optimization objectives, but in actual implementation, since the actual available capacity of the node is far lower than the rated value, it will inevitably lead to the inability to process materials in a timely manner after they arrive, causing a series of engineering infeasibility problems such as severe overload congestion, forced rejection, or illegal temporary storage. By introducing the variable of unavailable capacity, this application corrects the available carrying capacity of the target node from a static theoretical value to a dynamic value that reflects the current actual operating status, eliminates the spurious abundance of the solution space boundary, effectively prevents the problem of node overload allocation, and ensures the feasibility of the capacity carrying capacity of the optimization results.

[0064] Furthermore, dynamic availability is used as the upper limit of allocation constraints, including: the total amount of materials allocated to the target node shall not exceed its current dynamic availability.

[0065] That is, in the constraints of the optimization model, for any target node j, the sum of the total amount of all materials i to be allocated to it satisfies the following condition. The rigid constraints ensure that at the current time t, no material allocation will exceed the actual remaining load limit of the node.

[0066] In addition, the dynamic availability capability is updated dynamically over time.

[0067] Since the operating status of the target node evolves continuously over time, its currently allocated load... Its unusable capacity increases with the arrival of new materials and decreases with the disposal of old materials; Its dynamic availability will also increase or decrease as maintenance ends, production restriction policies are lifted, or new production restriction orders are issued. Therefore, dynamic availability... It has time-varying response characteristics to capture fluctuations in node load capacity in real time.

[0068] This application does not impose a single limitation on the specific update frequency and triggering method of dynamic availability capability; it can be flexibly set according to the data acquisition capabilities and scheduling requirements of the actual engineering system. Updates can be based on real-time monitoring data, i.e., through IoT sensors or online monitoring systems, triggering immediately upon detecting changes in facility load or shutdown signals. The recalculation and updating capabilities are suitable for dynamic scheduling scenarios with extremely high real-time requirements. It can also be updated based on rolling time windows, using historical operational data and predictive models to extrapolate the changing trends of facility load and unavailable capacity within a future time window, thus updating the dynamic availability capacity for subsequent periods in advance. This is suitable for scheduled scheduling scenarios requiring forward-looking planning. Furthermore, it can be updated in batches based on scheduling cycles, meaning that the latest status data of each node is collected centrally before the start of each daily, weekly, or scheduling batch, and all nodes are updated uniformly. The value is suitable for conventional planning scenarios where state changes are relatively gradual or scheduling frequency is low. Regardless of the triggering method used, the core is that dynamic availability can reflect the actual carrying status of the target node over time, rather than a static value that remains unchanged.

[0069] This embodiment reveals the formation mechanism of unavailable capacity and establishes a dynamic availability calculation and update mechanism, correcting the load-bearing capacity constraint of the target node from static virtual width to dynamic reality. This enables the optimization model to respond to changes in the node's operating status in real time, effectively avoiding overload allocation problems caused by ignoring the temporary unavailable status of facilities, and further improving the executability and stability of the optimization results in the actual dynamic operating environment of engineering projects.

[0070] In some embodiments, the optimization objective composition and the logic for determining the optimal allocation path in step S150 are further elaborated.

[0071] Specifically, the optimization objectives for solving multi-objective optimization problems include: minimizing total cost and minimizing environmental impact.

[0072] In practical resource allocation and route planning scenarios, economic cost and environmental impact are often two core and conflicting considerations. Minimizing total cost aims to reduce direct economic expenditures in transportation, processing, or disposal, while minimizing environmental impact aims to reduce carbon emissions, ecological damage, or potential environmental risks generated during system operation. For example, in hazardous waste disposal, choosing a nearby disposal facility can effectively reduce transportation costs, but this facility may lead to higher carbon emissions or environmental risks due to outdated technology. Conversely, choosing a distant facility with advanced technology and environmental friendliness can minimize environmental impact, but it will significantly increase transportation distance and costs. This inherent conflict between objectives makes traditional single-objective optimization methods unable to provide a reasonable solution that considers multiple demands. Multi-objective collaborative optimization mechanisms seek the optimal balance between two or more objectives, rather than simply pursuing the extreme optimality of a single objective.

[0073] It should be understood that although this embodiment uses the minimization of total cost and the minimization of environmental impact as typical optimization objectives, in other embodiments, depending on specific business requirements, the optimization objectives can be further extended to the minimization of total risk level, the maximization of resource utilization benefits, the minimization of transportation time, or the maximization of system stability, etc. As long as there are conflicting relationships between these objectives and they need to be coordinated and weighed, they can all be included in the multi-objective optimization function system of this application.

[0074] After solving the multi-objective optimization problem, the optimal allocation path is determined, including: obtaining the Pareto optimal solution set obtained from the multi-objective optimization problem; calculating the distance between each solution and the ideal solution and the negative ideal solution; calculating the relative proximity of each solution based on the distance; sorting the Pareto optimal solution set according to the relative proximity and selecting the optimal allocation path.

[0075] The Pareto optimal solution set consists of a set of multiple non-dominant feasible allocation schemes. An improvement in one objective by each scheme in the set inevitably leads to a deterioration in another objective. This presents decision-makers with a multi-dimensional group of conflicting schemes, rather than a single optimal solution. To transform this complex Pareto solution set into a single, directly executable decision in engineering, this application introduces a distance-based relative proximity ranking logic.

[0076] The specific calculation process of this sorting logic is as follows: First, obtain the Pareto optimal solution set from the multi-objective optimization output and construct a standardized decision matrix. Ideal solutions are virtual reference points composed of the optimal limits of each optimization objective; for example, the optimal limit of the total cost objective is the lowest cost among all options, and the optimal limit of the environmental impact objective is the lowest impact among all options. Negative ideal solutions are virtual reference points composed of the worst-case limits of each optimization objective; for example, the highest cost and the highest environmental impact.

[0077] Then, the distances between each solution and the ideal solution, as well as the negative ideal solution, are calculated. The distance from the i-th solution to the ideal solution is... The weighted Euclidean distance is usually used as the metric, and its calculation formula is as follows: ,in This represents the standardized value of the i-th scheme on the j-th evaluation index. Let represent the ideal optimal value of the j-th evaluation index; correspondingly, let represent the distance from the i-th solution to the negative ideal solution. The calculation formula is ,in This represents the negative ideal worst value of the j-th evaluation index.

[0078] Based on the above distances, the relative closeness of each solution is calculated. Its calculation formula is Relative closeness The value ranges from 0 to 1. The closer the value is to 1, the further the solution is from the negative ideal solution and the closer it is to the ideal solution, meaning the solution performs better overall under multi-dimensional conflict objectives; conversely, when... The closer the value is to 0, the worse the overall performance of the scheme.

[0079] Finally, according to relative closeness The Pareto optimal solution set is sorted in descending order, and the solution with the highest proximity is directly selected as the final optimal allocation path.

[0080] By using a ranking mechanism, the multidimensional conflict Pareto solution set, which was originally impossible to directly compare in terms of merit, is mapped onto a one-dimensional relative proximity scale, successfully transforming it into a ranking decision problem. This transformation not only provides decision-makers with an intuitive and quantitative basis for evaluating the merits of solutions, avoiding the dilemma of blindly selecting from numerous mutually exclusive solutions, but also significantly improves the applicability and decision-making efficiency of multi-objective optimization results in complex engineering environments.

[0081] It should be understood that this application does not impose a single limitation on the specific algorithm for solving multi-objective optimization problems or the specific evaluation method for sorting the solution set, but aims to establish a defensive depth based on the "distance-based proximity sorting idea". At the solution algorithm level, although this embodiment mentions the NSGA-III algorithm, multi-objective particle swarm optimization (MOPSO), decomposition-based multi-objective evolutionary algorithm (MOEA / D), or other suitable multi-objective optimization algorithms can also achieve the same function. The core is to be able to output a Pareto optimal solution set that satisfies the dual constraints.

[0082] At the ranking and evaluation level, although this embodiment elaborates on the TOPSIS ranking logic based on the distance between the ideal solution and the negative ideal solution, the Analytic Hierarchy Process (AHP), grey relational analysis, or other multi-index decision-making methods can also achieve comprehensive evaluation and ranking screening. As long as they can transform the Pareto solution set into a rankable one-dimensional decision index based on multi-dimensional target distance or correlation, they do not depart from the protection scope of this application.

[0083] As an example, to address the issues of missing data, abnormal fluctuations, and temporal misalignments in sludge production data from multiple industrial parks, a dynamic repair and anomaly correction system for industrial sludge production is constructed based on the coupling of process constraints, time-varying relationship graphs, and dynamic graph convolutional networks. The specific steps include: First, multi-source data generated by multiple wastewater treatment plants during actual operation were collected, including sludge production data, influent flow rate and pollutant concentration data, process operating parameters (such as sludge age, return ratio, aeration intensity, dissolved oxygen concentration, etc.), energy consumption data (such as electricity consumption, aeration energy consumption), and sludge moisture content.

[0084] The data sources include enterprise operation ledgers, online monitoring systems, DCS / SCADA systems, and environmental supervision platforms. The acquired data undergoes unified timestamp alignment, initial anomaly screening, unit unification, and coding processing to construct a standardized multi-source time-series dataset. Furthermore, to support subsequent dynamic graph construction and time-varying weight calculation, the system also extracts the data for each wastewater treatment plant within the sliding time window [t]. The operating load sequence and influent water quality sequence within [τ,t] are used, and basic reliability labels rs are assigned to different data sources, where s represents the data source category, which is used for subsequent attention allocation and credibility evaluation.

[0085] Based on the completed data acquisition, a typical process flow model was constructed for the key links in sludge generation during wastewater treatment, including primary sedimentation tank, biochemical reaction unit, secondary sedimentation tank and sludge treatment unit, to clarify the sludge generation pathway and its relationship with the pollutant removal process.

[0086] Based on this, and using the principles of material and energy balance, theoretical constraints on sludge production are established, including: the functional relationship between sludge production and COD removal, the reasonable range of sludge production per unit volume of water, and the correction relationship between sludge moisture content and statistical production, thus forming a system of process constraint rules for data verification.

[0087] Furthermore, to support subsequent dynamic fusion correction, this application assigns a constraint reliability coefficient cp to different process constraint rules, where p represents the process segment or constraint category to which the constraint belongs, and cp∈[0,1] represents the credibility of the constraint in the current scenario. The constraint reliability can be determined based on the fitting error of historical operating data, the stability of expert rules, or the completeness of process segment data.

[0088] Furthermore, multiple wastewater treatment plants are abstracted as nodes in a graph structure, and a time-varying relationship graph sequence that evolves over time is constructed: ;in, Represents the set of nodes of the wastewater treatment plant. Indicates time The set of edges, Indicates time The connection weight matrix is ​​used. Unlike static graphs with fixed weights, the nodes in this application... With nodes At any moment The association weight is defined as: ; in: Represents a node With nodes Spatial distance between them; Indicates process similarity; Indicates the similarity in processing scale; This indicates the correlation of operating load within the sliding time window; This indicates the similarity of influent water quality within the sliding time window; These are the weighting coefficients.

[0089] The operational load correlation can be expressed as: ; in, Represents a node At any moment Operating load indicators This represents the average value within the corresponding time window. The similarity of influent water quality can be expressed as: ; in, Represents a node At any moment The One influent water quality indicator, Indicates the number of water quality indicators. This is the normalization coefficient.

[0090] In this way, the connection between nodes is not only affected by static factors such as spatial distance, process similarity and scale difference, but also dynamically adjusted with changes in operating status, thus forming a dynamic graph sequence driven by operating status, providing a structural foundation for subsequent dynamic spatiotemporal modeling.

[0091] Based on the construction of the time-varying relationship diagram, the sludge production and related influencing factors of each wastewater treatment plant are organized according to the time series, forming a multi-dimensional data structure that includes time dimension, spatial node dimension, and feature dimension: Where N represents the number of nodes in the wastewater treatment plant, and F represents the feature dimension.

[0092] Based on this, characteristic variables related to sludge production were extracted, and a set of characteristic variables was constructed, the definitions of which are shown in Table 1.

[0093] Table 1

[0094] By uniformly arranging the aforementioned multi-source data, a spatiotemporal feature dataset is constructed for model input. .

[0095] Based on the constructed dynamic graph sequence and spatiotemporal feature data, a dynamic graph convolutional network model is introduced to model and predict sludge production data. Unlike ordinary ST-GCN which uses a fixed adjacency matrix, this application uses a time-varying weight matrix at each time step. Spatial feature aggregation is performed to achieve dynamic perception of changes in operational status. The spatial aggregation process can be represented as follows: ; in: Indicates the first Layer at time Node feature representation; ; This is the degree matrix; These are learnable parameters; The activation function is used. In the time dimension, gated sequential units are employed. A model was constructed to extract the dynamic pattern of sludge production over time. The model outputs predicted sludge production values. and forecast uncertainty .

[0096] The prediction uncertainty can be determined by the model's multiple sampling results, the ensemble prediction variance, or the confidence interval width. After the model training is complete, the actual observed values ​​will be used... Compared with model predictions By comparison, a standardized outlier bias is constructed: ;in, To prevent small constants with a denominator of zero. When Exceeding the preset threshold When the data is abnormal, it will be identified as abnormal data, thus enabling automatic identification of sudden changes, abnormal fluctuations, and unreasonable data points.

[0097] After anomaly identification is completed, the model prediction results are fused with process constraint rules to perform secondary correction on the abnormal data.

[0098] Specifically, the prediction results of the dynamic graph convolutional network Compared with the theoretical value calculated based on process constraints Weighted fusion is performed to obtain the final correction result: ; in, The fusion weights, which are dynamically adjusted over time, are no longer fixed values, but are determined jointly based on model prediction uncertainty and process constraint reliability. Their expression can be defined as: ; in: This indicates the uncertainty in the model's predictions; The reliability of process constraints at the current moment can be represented by the reliability coefficient of the corresponding process segment constraints. Combined to obtain; To adjust the parameters.

[0099] When the uncertainty of model predictions increases When the reliability of process constraints is reduced, the system tends to adopt the results of process constraints more readily; when the reliability of process constraints is low... As the data volume increases, the system tends to adopt data-driven prediction results. Through the above dynamic adjustment strategy, the corrected data simultaneously satisfies statistical regularities and engineering mechanism constraints, and adaptively balances the two types of information sources according to different scenarios.

[0100] To address the issues of missing and time misalignment in sludge production data, this application introduces a multi-source spatiotemporal attention mechanism for sludge production scenarios, which weights and fuses information from different time points, spatially adjacent nodes, and data sources.

[0101] For nodes At any moment For the missing values, construct three attention weights: Temporal attention weights are used to characterize the contribution of nearby historical moments to the current missing value: ; in, For the time neighborhood set, It can be determined by time distance, similarity of historical trends, and prediction residuals.

[0102] Spatial attention weights are used to characterize the contribution of associated plants to the current missing value: ; in, Indicates time Next and Node The set of connected neighboring nodes. Weights from dynamic graphs The similarity of node features and the degree of historical collaboration determine this.

[0103] Data source attention weights are used to characterize the credibility contribution of different data sources in the current scenario: ; in, Represents a collection of data sources. Based on data source reliability tags The current consistency error and historical stability are jointly determined.

[0104] Based on the above three types of weights, missing values ​​are jointly imputed: ; in, The fusion coefficients for the three complementary components of time, space, and data source satisfy... . For time-misaligned data, local time-shift correction is performed on the observation sequence with the objective of minimizing the prediction residuals between adjacent time periods, and the optimal shift is selected. : ; This enables the correction of data that is lagging or ahead of time, restoring the consistency of multi-source data in the spatiotemporal dimension.

[0105] After data correction and completion, the processing results are evaluated to establish a reliability evaluation index system for sludge production data. The evaluation index comprehensively considers: the degree of data deviation; the magnitude of correction; the degree to which process constraints are met; spatial consistency; and the intensity of dependency on data completion.

[0106] The evaluation indicators are defined as follows: , representing the forecast consistency index; , indicating the calibration stability index; , indicating the degree of satisfaction of process constraints; , representing the spatial consistency index; This represents an indicator of the completeness or dependency level of the original data. Further, a comprehensive credibility score is constructed: ; in, Let the weighting coefficients satisfy: ; Based on overall credibility score Divide data into levels, for example: Highly reliable; Highly credible; Generally credible; Low credibility / high risk.

[0107] The final output includes data on the sludge production after treatment. Missing value completion results Data quality labels, credibility levels, and abnormal risk warnings provide reliable data support and decision-making basis for regulatory authorities or operation management departments.

[0108] like Figure 2 As shown, this embodiment provides a device for determining the disposal path of fly ash. This device and the method embodiment belong to the same technical concept. Through the division of logical modules and data flow, it achieves a functional mapping that completely corresponds to the method steps, aiming to ensure compliance of the object while solidifying the core mechanism of the aforementioned method into an executable virtual system architecture. The device mainly includes an acquisition module 201, a network construction module 202, an adaptability calculation module 203, a dynamic capability determination module 204, and an optimization solution module 205.

[0109] The acquisition module 201 is used to acquire the attribute parameters of the material to be allocated, the process requirement parameters of the target node, the rated capacity, the currently allocated load, and the unavailable capacity.

[0110] Specifically, the acquisition module 201 is the data entry point and driving source for the entire device. It is responsible for extracting, cleaning, and formatting various basic parameters from external data sources or human-machine interfaces. In terms of data flow logic, the acquisition module 201 outputs attribute parameters and process requirement parameters to the adaptability calculation module 203, outputs rated capacity, currently allocated load, and unavailable capacity to the dynamic capacity determination module 204, and simultaneously outputs all basic data to the network construction module 202 as material for topology modeling.

[0111] It should be understood that although the acquisition module 201 is described as a unified logical unit in this embodiment, in other embodiments, the acquisition module 201 may be further subdivided into attribute parameter acquisition submodule, process requirement acquisition submodule and capability status acquisition submodule according to the heterogeneity of data sources, as long as it can complete the aggregation and distribution of all input data.

[0112] Network construction module 202 is used to construct a flow network for materials to be allocated to the target node.

[0113] Specifically, the network construction module 202 receives the basic data output by the acquisition module 201, abstracts each source and target node into network nodes, and abstracts the feasible transportation paths between the source and target into edges. It then assigns attribute parameters such as distance, unit transportation cost, unit transportation carbon emissions, and transportation risk coefficient to the edges, thereby generating a flow network model expressed in a graph structure. This module outputs the constructed flow network topology data to the optimization solution module 205, serving as the environmental foundation and topology carrier for subsequent multi-objective optimization.

[0114] The adaptability calculation module 203 is used to calculate the adaptability between the material to be allocated and the target node based on the attribute parameters and the process requirement parameters, and to use the adaptability as a feasibility constraint to eliminate infeasible paths.

[0115] Specifically, the fitness calculation module 203 receives the attribute parameters and process requirement parameters output by the acquisition module 201, executes the fitness quantification calculation logic, and determines the feasibility of path branches in the flow network based on the calculation results. For paths determined to be infeasible, this module directly sets their decision variables to zero or removes them from the candidate edge set. This module outputs the set of effective paths after feasibility constraint pruning and the fitness quantification values ​​of each effective path to the optimization solution module 205 as the first line of defense data for correcting the solution space boundary.

[0116] The dynamic capacity determination module 204 is used to determine the dynamic availability capacity of the target node based on the rated capacity, the currently allocated load and the unavailable capacity, and to use the dynamic availability capacity as the upper limit of the allocation constraint.

[0117] Specifically, the dynamic capacity determination module 204 receives the rated capacity, currently allocated load, and unavailable capacity output by the acquisition module 201, executes the dynamic availability capacity calculation logic, and corrects the static theoretical rated value to a dynamic value that reflects the current actual operating state. This dynamic value is then used as the rigid upper limit of subsequent allocation constraints. This module outputs the dynamic availability capacity constraint upper limit data for each target node to the optimization solution module 205, serving as the second line of defense data for correcting the solution space boundary.

[0118] The optimization solution module 205 is used to perform multi-objective optimization on the flow network based on the feasibility constraints and the allocation constraints to determine the optimal allocation path.

[0119] Specifically, the optimization solution module 205 is the decision-making center of the entire device. It simultaneously receives the flow network topology data output by the network construction module 202, the feasibility constraint data output by the fitness calculation module 203, and the allocation constraint upper limit data output by the dynamic capability determination module 204. Under the joint constraints of these dual constraint data, the optimization solution module 205 initiates a multi-objective optimization algorithm within the corrected effective solution space boundary, and finally outputs the optimal allocation path scheme that takes into account both process matching and capacity carrying capacity feasibility.

[0120] Through the coordinated operation and data flow of the aforementioned logical modules, the fly ash disposal path determination device of this embodiment solidifies the core mechanism of synchronously injecting adaptation constraints and dynamic availability constraints to correct the solution space boundary in the aforementioned method into the input-output dependency relationship between modules. This achieves dual feasibility of path allocation in terms of process matching and capacity carrying capacity, significantly improving the executability and stability of the optimization results in the actual dynamic operation environment of engineering. It should be understood that the modules described in this embodiment refer to logical function modules. In actual software or hardware implementation, multiple modules can be merged into one entity module, or one module can be split into multiple sub-modules, as long as they can complete the logical functions and data flow defined above.

[0121] In some embodiments, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the fly ash disposal path determination method described in the foregoing embodiments of this application.

[0122] Specifically, the computer-readable storage medium serves as the physical carrier of the software code, and the computer program logic stored within it completely encapsulates the core mechanism of this application. When the program is loaded and executed by a processor in various computing devices, the processor will sequentially perform the following steps: obtaining the attribute parameters of the material to be allocated, the process requirement parameters of the target node, the rated capacity, the currently allocated load, and the unavailable capacity; constructing a flow network from the material to be allocated to the target node; calculating the fit between the material to be allocated and the target node based on the attribute parameters and process requirement parameters, and using the fit as a feasibility constraint to eliminate infeasible paths; determining the dynamic available capacity of the target node based on the rated capacity, the currently allocated load, and the unavailable capacity, and using the dynamic available capacity as the upper limit of the allocation amount constraint; and performing multi-objective optimization on the flow network based on the feasibility constraint and the allocation amount constraint to determine the optimal allocation path. By executing program instructions one by one, the processor, which originally only had general computing capabilities, is temporarily configured as a dedicated logic operation unit with path optimization decision-making capabilities. This allows for the simultaneous injection of material and process adaptation feasibility constraints and node dynamic capability allocation constraints into the multi-node allocation optimization model, correcting the solution space boundary of the optimization model and ensuring that the final output optimal allocation path is feasible in both process matching and capability carrying capacity.

[0123] It should be understood that the computer-readable storage medium referred to in this application can be any physical or logical medium capable of containing or storing program code, whose core function is to allow an instruction execution system to read and execute program instructions. This includes not only traditional magnetic storage media such as portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic tapes, or floppy disks, but also optical storage media such as optical discs (CD-ROM, DVD, etc.), and semiconductor storage media such as flash memory and solid-state drives (SSDs). Even in some distributed or cloud computing scenarios, the medium can be a logically virtualized storage space. As long as it can stably carry and provide the computer program code of this application, enabling the processor to read and execute this code through a bus or network interface to implement the method steps of this application, it should fall within the protection scope of the computer-readable storage medium of this application.

[0124] The above description is merely an illustrative description of the computer-readable storage medium embodiments of this application, and not a limiting definition. Those skilled in the art should understand that with the evolution of storage technology, any new types of storage media that may emerge in the future, as long as they can achieve the same function as this embodiment—that is, storing and allowing the processor to execute the path optimization program of this application—can be considered equivalent replacements for this embodiment.

[0125] In some embodiments, Figure 3This application provides a hazardous waste disposal pathway optimization system. System 100 is the physical embodiment of the aforementioned method embodiments at the product level, designed to cover the original application scenario of hazardous waste disposal. Through the coordinated operation of hardware and software, it solidifies the dual defense mechanism of adaptability constraints and dynamic availability constraints into a deployable engineering system.

[0126] The system includes a processor 30, a memory 31, a bus 32, and a communication interface 33. The processor 30, the communication interface 33, and the memory 31 are connected via the bus 32. The processor 30 is used to execute executable modules, such as computer programs, stored in the memory 31.

[0127] The memory 31 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 33 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc.

[0128] Bus 32 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 3 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0129] The memory 31 is used to store programs. After receiving an execution instruction, the processor 30 executes the programs. The methods executed by the apparatus for stream process definition disclosed in any of the foregoing embodiments of this application can be applied to the processor 30 or implemented by the processor 30.

[0130] The processor 30 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 30 or by instructions in software form. The processor 30 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 31, and the processor 30 reads the information from memory 31 and, in conjunction with its hardware, completes the steps of the above method.

[0131] The computer program product of the readable storage medium provided in the embodiments of this application includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the foregoing method embodiments. For specific implementation, please refer to the foregoing method embodiments, which will not be repeated here.

[0132] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0133] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The scope of protection of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application. Such modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for determining the disposal path of fly ash, characterized in that, include: Obtain the attribute parameters of the materials to be allocated, the process requirements parameters of the target node, the rated capacity, the currently allocated load, and the unavailable capacity; Construct the flow network of the materials to be allocated to the target nodes; Based on the attribute parameters and the process requirement parameters, the fit between the material to be allocated and the target node is calculated, and the fit is used as a feasibility constraint to eliminate infeasible paths. Based on the rated capacity, the currently allocated load, and the unavailable capacity, the dynamic availability capacity of the target node is determined, and the dynamic availability capacity is used as the upper limit of the allocation constraint. Based on the feasibility constraints and the allocation constraints, the flow network is optimized using a multi-objective approach to determine the optimal allocation path.

2. The method for determining the fly ash disposal path according to claim 1, characterized in that, The adaptability, as a feasibility constraint, is used to eliminate infeasible paths, including: Hard constraint: Boolean judgment based on qualification or category; if not satisfied, the path is disabled. Soft constraint: The path is disabled if the degree of matching between the attribute parameters and the process requirement parameters is less than the preset threshold.

3. The method for determining the fly ash disposal path according to claim 2, characterized in that, The calculation of the quantized matching degree of the soft constraint includes: Calculate the degree of individual matching between the material to be allocated and the target node on each indicator; The overall matching degree is obtained by weighted summation of the matching degrees of each individual item; The degree of matching of each indicator is determined based on the correspondence between the attribute parameters and the process requirement parameters on that indicator, and the weighting coefficient reflects the importance of each indicator in the comprehensive evaluation.

4. The method for determining the fly ash disposal path according to claim 1, characterized in that, Determining the dynamic availability of the target node based on the rated capacity, the currently allocated load, and the unavailable capacity includes: Dynamic availability equals rated capacity minus currently allocated load minus unavailable capacity; The unavailable capacity includes capacity occupied due to maintenance, production restrictions, or policy limitations.

5. The method for determining the fly ash disposal path according to claim 4, characterized in that, The step of using the dynamically available capacity as the upper limit of the allocation constraint includes: The total amount of materials allocated to the target node shall not exceed its current dynamic available capacity; The dynamic availability capability is updated dynamically over time.

6. The method for determining the fly ash disposal path according to claim 1, characterized in that, The optimization objectives of the multi-objective optimization solution include: Minimize total cost; Minimize environmental impact.

7. The method for determining the fly ash disposal path according to claim 1, characterized in that, Determining the optimal allocation path includes: Obtain the Pareto optimal solution set obtained from multi-objective optimization; Calculate the distance between each solution and the ideal solution and the negative ideal solution; The relative proximity of each solution is calculated based on the distance. The Pareto optimal solution set is sorted according to relative proximity, and the optimal allocation path is selected.

8. A device for determining the disposal path of fly ash, characterized in that, include: The acquisition module is used to acquire the attribute parameters of the materials to be allocated, the process requirement parameters of the target node, the rated capacity, the currently allocated load, and the unavailable capacity. A network construction module is used to construct the flow network of the materials to be allocated to the target nodes; The adaptability calculation module is used to calculate the adaptability between the material to be allocated and the target node based on the attribute parameters and the process requirement parameters, and to use the adaptability as a feasibility constraint to eliminate infeasible paths. The dynamic capacity determination module is used to determine the dynamic availability capacity of the target node based on the rated capacity, the currently allocated load, and the unavailable capacity, and to use the dynamic availability capacity as the upper limit of the allocation constraint. The optimization solution module is used to perform multi-objective optimization on the flow network based on the feasibility constraints and the allocation constraints to determine the optimal allocation path.

9. A computer-readable storage medium, characterized in that, It stores a computer program thereon, which, when executed by a processor, implements the method as described in any one of claims 1 to 7.

10. A system for determining the disposal path of fly ash, characterized in that, Includes a processor configured to perform the method as described in any one of claims 1 to 7.

Images