A credit-driven trusted collaborative scheduling method and system for watershed sudden water pollution events

Abstract

The application discloses a credit-driven credible collaborative scheduling method and system for watershed sudden water pollution events, and the method comprises the following steps: constructing a watershed management object graph; calculating a scheduling credibility level according to multi-source sensing data; screening event-related facilities based on pollution influence range, facility action range, scheduling authority and available state; constructing a scheduling capability portrait for the related facilities; generating a candidate scheme by contract network negotiation based on the capability portrait and credit value, and evaluating the scheme effect by counterfactual simulation; checking the authority, signature, facility state and cross-regional collaborative commitment through a blockchain smart contract, and generating an on-chain scheduling voucher; verifying the receipt, calculating the effect deviation and updating the credit value after execution, and the credit value inversely affects the subsequent negotiation weight and scheduling qualification. The application realizes a closed-loop credible scheduling of "grading-negotiation-verification-execution-feedback", and improves the executability, credibility and traceability of cross-regional collaborative management.

Description

Technical Field

[0001] This invention relates to the field of emergency management and intelligent scheduling technology for watershed water environment, specifically to a collaborative management and scheduling method, system, and equipment for sudden water pollution events in watersheds based on facility capability profiling and blockchain smart contracts. This application combines watershed management object mapping, multi-agent contract network negotiation, counterfactual effect assessment, and blockchain smart contract technology to achieve closed-loop processing in sudden water pollution events, including object selection, generation of reliable collaborative solutions, issuance of on-chain credentials, and execution credit feedback. Background Technology

[0002] Sudden water pollution incidents in river basins are characterized by rapid spread, cross-regional nature, multiple responsible parties, and complex facility types. Emergency response urgently requires coordination among various entities, including sluice gates, pumping stations, sewage treatment facilities, water storage facilities, water intakes, polluting enterprises, and local management units. Existing dispatch methods largely rely on a central platform or manual consultations to generate response plans, but these methods suffer from the following technical shortcomings: The scheduling capacity of facilities lacks dynamic characterization. Information such as gate opening, pumping station capacity, remaining facility processing capacity, reservoir capacity, water intake boundaries, and historical response reliability are dynamically changing. Existing methods do not incorporate these into a computable scheduling capacity description framework, leading to a mismatch between theoretical solutions and on-site executability. Existing knowledge graph-based water diversion methods focus primarily on data association queries and anomaly attribution, failing to uniformly model action boundaries, constraints, and credit.

[0003] The lack of reliable records regarding authority relationships and collaborative commitments is a significant issue. Cross-regional dispatching involves multiple levels of entities and various facilities, but existing systems struggle to prove in advance the attribution of dispatching authority, the validity of collaborative commitments, and the consistency of execution. While there are studies on the application of blockchain in water environment monitoring and evidence storage, as well as multi-agent collaborative early warning, these studies only address the reliable authorization of dispatching instructions, the solidification of collaborative commitments, and the issuance of credentials, lacking authority verification and non-repudiation execution mechanisms for governance-oriented dispatching.

[0004] The proposed solutions lack pre-implementation verification and a closed-loop feedback mechanism. Existing solutions are typically issued directly after generation, lacking forward-looking effect assessments and constraint verification based on hydrodynamic-water quality models or counterfactual simulations. This leads to unfeasible or inefficient solutions potentially entering the actual implementation phase. Furthermore, there is a lack of quantitative comparison of effect deviations and systematic feedback mechanisms after implementation. Although some studies use credit for post-implementation evaluation, it is not considered a dynamic input for real-time impact on facility selection, negotiation weighting, and scheduling eligibility.

[0005] The various technological elements have not yet formed a deeply integrated closed-loop system. Elements such as knowledge graphs, multi-agent systems, counterfactual simulation, and blockchain are mostly used independently or loosely coupled in existing technologies, and have not yet formed a complete closed loop from event classification, facility screening, solution screening, solution negotiation, forward-looking assessment, trusted execution, and credit feedback.

[0006] Therefore, there is a need for a watershed emergency water pollution event governance scheduling method that can deeply couple the governance object map, facility capacity profile, multi-agent negotiation, counterfactual effect assessment, and blockchain trusted execution. Summary of the Invention

[0007] The purpose of this invention is to provide a credit-driven trusted collaborative scheduling method and system for sudden water pollution events in watersheds, addressing the aforementioned problems.

[0008] The technical solution of the present invention is as follows: A credit-driven, trustworthy collaborative scheduling method for sudden water pollution events in a watershed includes the following steps: Construct a map of watershed governance objects, which includes river section nodes, monitoring section nodes, sewage outlet nodes, gate nodes, pumping station nodes, sewage treatment facility nodes, water storage facility nodes, water intake nodes, administrative region nodes, responsible entity nodes, as well as hydrological connectivity edges, pollution propagation edges, facility function edges, dispatch authority edges, local management edges, and data reporting edges. Acquire multi-source sensing data on sudden water pollution events, and calculate scheduling reliability scores based on pollutant abnormal intensity, abnormal duration, consistency of hydrological propagation, risk of sensitive targets, data integrity, and manual verification status. The pollution impact range is calculated based on the location of the event and the hydrological connectivity edges in the watershed management object map. The set of facilities associated with the event is then selected based on whether there is an intersection between the pollution impact range and the facility's operational range, the scheduling authority determination results, and the facility availability status. For each facility in the event-related facility set, construct a scheduling capability profile. The scheduling capability profile shall include at least the set of executable actions, action parameter boundaries, response time, security constraints, ecological constraints, and historical execution credit value. Based on scheduling capability profiles and historical execution credit scores, a contract network negotiation mechanism is used to generate candidate governance scheduling schemes. Conduct counterfactual effect assessments on candidate governance scheduling schemes, and reject, revise, or trigger renegotiation based on the assessment results; Candidate solutions that pass the counterfactual effect evaluation are submitted to the blockchain smart contract for verification. Once the verification is passed, an on-chain scheduling certificate is generated. Obtain execution receipts and multi-source feedback data after facility execution, verify the consistency between execution receipts and on-chain scheduling credentials, calculate governance effect deviation, update historical execution credit values ​​based on governance effect deviation, and write the updated historical execution credit values ​​back to the scheduling capability profile.

[0009] Furthermore, the step of updating the historical execution credit value based on the governance effect deviation and writing the updated historical execution credit value back to the scheduling capability profile also includes: The updated historical execution credit value has a reverse impact on the bidding score weight of the corresponding facility in subsequent contract network negotiations; when the updated historical execution credit value is lower than the preset freezing threshold, the corresponding facility will be removed from the set of automatically dispatched facilities and marked as a candidate facility for manual review only; when the updated historical execution credit value is higher than the preset reward threshold, the bidding score weight of the corresponding facility will be increased in subsequent negotiations. The scheduling trust score is input into the blockchain smart contract as a dynamic parameter. When the scheduling trust score is in the preset medium range, the smart contract automatically adds additional cross-regional subject signature verification conditions.

[0010] Furthermore, the counterfactual effect evaluation of candidate governance scheduling schemes, and the rejection, modification, or triggering of renegotiation based on the evaluation results, specifically includes: Construct counterfactual scenarios corresponding to candidate governance scheduling schemes, and predict the peak pollution reduction rate and the arrival delay time of sensitive target pollution after implementing the scheme through hydrodynamic water quality models or pollution diffusion proxy models. When the counterfactual effect score is lower than the preset threshold or the scheme violates safety or ecological constraints, if there are correctable parameters, the facility action parameters in the candidate scheme are adjusted according to the action parameter boundary and re-evaluated; if there are no feasible corrections, the specific problems identified in the evaluation are encoded as correction instructions and returned to the contract network negotiation mechanism, the credit score weight of the corresponding facility in the next round of negotiation is reduced, and it is forced to provide a corrected contribution estimate when re-bidding.

[0011] Furthermore, the calculation of the scheduling reliability score based on the abnormal intensity and duration of pollutants, consistency of hydrological propagation, risk of sensitive targets, data integrity, and manual verification status specifically includes: Calculate the scheduling reliability score using the following formula: , The abnormal intensity of pollutants is calculated as follows: , in, For actual measured concentration, Background concentration, These are the standard limits for the corresponding pollutants; The duration of the anomaly is calculated as follows: , in, The duration of continuous anomalies, The preset duration threshold; Hydrological propagation consistency is calculated as follows: , in, The time when the anomaly was observed at the downstream monitoring point. The arrival time is predicted based on upstream anomalies and propagation delays. This is the allowable deviation threshold; When the scheduling trust score is lower than the first threshold, only a simulated scheduling suggestion is generated and no on-chain scheduling certificate is issued; when the scheduling trust score is between the first threshold and the second threshold, a candidate scheme is generated and enters the manual review process; when the scheduling trust score is not lower than the second threshold, the process of smart contract verification and certificate issuance is allowed.

[0012] Furthermore, the process of verifying the consistency between the execution receipt and the on-chain scheduling certificate, calculating the governance effect deviation, and updating the historical execution credit value based on the governance effect deviation specifically includes: Verify the consistency of operation of each facility as follows: , in, This indicates that the actual action is consistent with the action on the voucher. This indicates that the execution time satisfies the voucher constraint. This indicates that the signature of the executing entity is valid; Calculate the treatment effect deviation using the following formula:

[0013] in, To predict the pollution reduction rate, To measure the actual pollution reduction rate, To predict response time, This is the actual response time. For the degree of actual violation of constraints, For the sake of completeness of the receipt; Update historical execution credit scores using the following formula: , in, To preserve the historical coefficient, Assess the responsibility weight of the facility in the plan. Penalties for failure to execute according to the voucher, execution exceeding the time limit, or false receipts. This means that the credit score will be limited to between 0 and 1.

[0014] Furthermore, the generation of candidate governance scheduling schemes based on scheduling capability profiles and historical execution credit values ​​using a contract network negotiation mechanism specifically includes: The event scheduling agent publishes task announcements to the event-related facility set. Each facility agent generates bidding information based on its own scheduling capability profile. The bidding information includes the proposed action and parameters, expected governance contribution, expected response time, action cost, security or ecological risks, and historical execution credit value. The negotiation score for each bid is calculated using the following formula: , in, To the expected contribution to governance, To estimate the response time, As a cost of action, For safety or ecological risks, For historical execution credit scores, to As weight; The bids are ranked based on the negotiated scores, generating the top-K candidate solutions or a set of Pareto candidate solutions that meet the constraints.

[0015] Furthermore, after generating the candidate governance scheduling scheme, the process also includes: A conflict constraint matrix is ​​used to determine conflicts between candidate facility actions. Indication facilities and facilities The corresponding actions cannot be selected simultaneously; when a conflict is determined, a weighted arbitration is conducted based on the risk of sensitive targets, scheduling credibility score, historical execution credit value and expected governance contribution. If the conflict cannot be resolved after arbitration, renegotiation is triggered, requiring both parties to submit revised bids.

[0016] Furthermore, the initial value of the historical execution credit value in the scheduling capability profile is determined by statistical analysis of historical scheduling records or set as a default value. The action parameter boundaries include the minimum value, maximum value, adjustment step size, and maximum rate of change of the action parameters. The safety constraints include upstream water level warning constraints, and the ecological constraints include downstream ecological flow lower limit constraints.

[0017] Furthermore, the process of filtering the set of facilities associated with the event based on whether there is an overlap between the pollution impact range and the effective range of each treatment facility, the scheduling authority determination result, and the facility availability status specifically includes: The facility will be determined according to the following criteria. Add to event-related facility collection : , in, The extent of the pollution impact, For the area of ​​effect of the facility, This refers to the permission determination result obtained by querying based on scheduling permissions. For facilities at all times The availability status; The availability status of a facility is determined by a combination of factors, including its operational status, fault status, maintenance status, manual lockout status, communication status, and security boundaries. If any of these statuses indicate that a facility is unavailable, it is excluded from the set of facilities associated with the event.

[0018] This application also includes a credit-driven trusted collaborative scheduling system for sudden water pollution events in a watershed, and implements a credit-driven trusted collaborative scheduling method for sudden water pollution events in a watershed, including: The watershed governance object map construction module is used to construct a watershed governance object map that includes river sections, monitoring sections, sewage outlets, gates, pumping stations, sewage treatment facilities, storage facilities, water intakes, administrative regions, responsible entities, and relationships such as hydrological connectivity, pollution propagation, facility functions, and scheduling authority. The scheduling credibility level calculation module is used to calculate the scheduling credibility score based on the abnormal intensity of pollutants, the duration of abnormality, the consistency of hydrological propagation, the risk of sensitive targets, data integrity, and the status of manual verification. The event-related facility filtering module is used to filter the set of event-related facilities based on the scope of pollution impact, the scope of facility operation, scheduling authority relationships, and facility availability status. The facility scheduling capability profile building module is used to build a scheduling capability profile for each facility in the event-related facility set, including a set of executable actions, action parameter boundaries, response time, security constraints, ecological constraints, and historical execution credit value. The contract negotiation module is used to generate candidate governance scheduling schemes based on scheduling capability profiles and historical execution credit values, through facility intelligent agents bidding and negotiation scoring ranking. The counterfactual assessment module is used to predict the effects and verify the constraints of candidate governance scheduling schemes, and to reject or modify candidate schemes or trigger the contract network negotiation module to renegotiate based on the assessment results. The blockchain smart contract module is used to verify the scheduling authority, responsible entity signature, facility status, action parameter boundaries, and cross-regional collaborative commitment of candidate solutions. After the verification is passed, an on-chain scheduling certificate is generated. The on-chain scheduling certificate includes at least the solution hash, multi-party signatures, validity period, and execution conditions. The execution verification and credit update module is used to obtain execution receipts and multi-source feedback data, verify the consistency between execution receipts and on-chain scheduling credentials, calculate governance effect deviation and update historical execution credit values ​​based on governance effect deviation, and write the updated historical execution credit values ​​back to the facility scheduling capability profile building module. The system is configured as follows: when the updated historical execution credit value is lower than the preset freezing threshold, the event-related facility screening module excludes the facility from the set of automatically scheduled facilities; the scheduling trust score is input as a dynamic parameter into the blockchain smart contract module to dynamically adjust the stringency of the contract verification rules.

[0019] Compared with existing technologies, the advantages of this invention are: 1. By using a watershed governance object map, hydrological connectivity, facility scope of action, and scheduling authority relationships are uniformly modeled into a computable directed heterogeneous graph. This allows for the quantitative determination of event-related facilities based on pollution propagation paths, facility scope of action, authority status, and availability, overcoming the arbitrariness and incompleteness caused by traditional methods relying on manual experience for facility selection. Simultaneously, by constructing a scheduling capability profile for each facility, including credible actions, action parameter boundaries, response time, safety constraints, ecological constraints, and historical execution credit values, the actual adjustability of the facility is structurally expressed before execution, effectively avoiding the technical deficiency of mismatch between theoretical scheduling schemes and on-site executable capabilities.

[0020] 2. By employing a contract network negotiation mechanism based on facility capability profiles and credit scores, each facility agent can independently bid based on its own capabilities, constraints, and credit. The negotiation scoring function then ranks, optimizes, and arbitrates conflicts, addressing the problems of lack of agreements for multi-type facility collaboration, lack of conflict arbitration, and absence of credit participation in real-time scheduling decisions in existing methods. Specifically, the conflict constraint matrix formalizes and automatically arbitrates the mutual exclusion relationships between facility actions; the negotiation scoring function uses credit scores as a weighting factor that influences the bidding ranking in real time, causing low-credit facilities to be downgraded or frozen during the negotiation phase, thus improving the actual executability of candidate solutions.

[0021] 3. This application introduces a counterfactual effect evaluation mechanism before on-chain issuance. It uses a hydrodynamic water quality model or proxy model to quantitatively predict the pollution reduction effect, response time, action cost, safety risk, and constraint violation degree of candidate solutions, and sets an effect scoring threshold as an issuance gate. When a candidate solution fails to meet constraints or has insufficient effect, the counterfactual evaluation module can return the reasons for the failure and correctable parameters, driving the system to correct action parameters or trigger contract network renegotiation. This avoids the risk of unexecutable or insufficiently effective solutions being directly issued and executed. This mechanism transforms post-event remediation into pre-event prevention, significantly improving the scientific nature and reliability of the scheduling scheme.

[0022] 4. By binding scheduling permission verification, multi-party collaborative commitments, facility status oracle data, action parameter boundaries, cross-regional collaborative rules, and counterfactual assessment result hashes into non-repudiable on-chain scheduling credentials through blockchain smart contracts, the issuance, commitment, execution, and accountability of scheduling instructions have a unified and credible basis. After execution, the governance effect deviation is calculated and the facility's historical execution credit value is updated by comparing the receipt verification with multi-source feedback data. The updated credit value inversely affects the facility selection, negotiation weight, and scheduling eligibility in subsequent events, realizing dynamic closed-loop control of "execution performance - credit evaluation - scheduling permission", overcoming the shortcomings of existing systems that have records but no feedback and assessments but no real-time control. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the overall process of the method described in this application.

[0024] Figure 2 This is a schematic diagram of the data structure of the watershed management object map in this application.

[0025] Figure 3 This is a schematic diagram of the multi-agent contract network negotiation process based on capability profiles in this application.

[0026] Figure 4 This is a schematic diagram of the smart contract scheduling certificate generation and execution receipt verification process in this application.

[0027] Figure 5 This is a schematic diagram illustrating the closed loop of governance effect deviation calculation and credit value update in this application.

[0028] Figure 6 This is a schematic diagram illustrating the specific implementation scenarios and data sources for this application.

[0029] Figure 7 This is a schematic diagram of the closed loop of gate control before issuance and feedback after execution of this application. Detailed Implementation

[0030] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0031] The features and performance of the present invention will be further described in detail below with reference to embodiments.

[0032] Please see Figure 1-7 A credit-driven, trustworthy collaborative scheduling method for sudden water pollution events in a watershed, such as... Figure 1 As shown, it includes the following steps: like Figure 2 As shown, a map of watershed governance objects is constructed: The objects of watershed management are represented as directed heterogeneous graphs: , in, The set of nodes includes at least river section nodes, monitoring section nodes, sewage outlet nodes, gate nodes, pumping station nodes, sewage treatment facility nodes, storage facility nodes, water intake nodes, administrative region nodes, and responsible entity nodes. The set of edges includes at least the hydrological connectivity edge, the pollution propagation edge, the facility function edge, the dispatch authority edge, the local management edge, and the data reporting edge. A collection of attributes representing a node or edge.

[0033] Each node The attributes should at least include object type, spatial location, administrative region, associated entity, status timestamp, and data source. Each hydrological connectivity edge The attributes include at least flow direction and distance. Average flow velocity Propagation delay and pollution attenuation coefficient The propagation delay can be expressed as: , in, The minimum positive number to prevent calculation errors caused by a flow rate of zero.

[0034] The dispatch authority edge indicates whether a regulatory or responsible entity has the authority to issue dispatch instructions to a facility, confirm execution results, or sign collaborative commitments. Through this graph, the system can quickly query pollution propagation paths, sensitive targets, candidate facilities, responsible entities, and dispatch authority after an event occurs, providing a topological basis and authority determination criteria for subsequent facility screening.

[0035] This step provides the data infrastructure for the entire method. Existing knowledge graphs are mostly used for data association queries and anomaly attribution. By introducing scheduling permission edges and facility action edges, the graph can not only support information retrieval, but also directly support permission determination and facility influence range calculation, providing a topology calculation basis for subsequent facility screening.

[0036] Calculate the scheduling reliability level for emergencies: Acquire multi-source sensing data related to sudden water pollution events, including at least water quality monitoring data, hydrological data, online monitoring data of discharge outlets, facility status data, image or video recognition data, remote sensing data, and manually verified data. (This is in response to the event.) Calculate the credibility score: , in, Indicates the abnormal intensity of pollutants. Indicates the duration of the anomaly. Indicates consistency in hydrological propagation. Indicates the risk of sensitive targets. Indicates data integrity. Indicates manual verification status; to Let be the weight, and satisfy... The specific calculation methods for each component are as follows: The abnormal intensity of pollutants can be calculated as follows: , in, For actual measured concentration, Background concentration, These are the standard limits for the corresponding pollutants.

[0037] The duration of the anomaly can be expressed as: , in, The duration of continuous anomalies, This is a preset duration threshold.

[0038] Hydrological propagation consistency is used to determine whether the temporal and spatial relationships of anomalies conform to the laws of water flow propagation, and can be expressed as: , in, The time when the anomaly was observed at the downstream monitoring point. The arrival time is predicted based on upstream anomalies and propagation delays. This is the allowable deviation threshold.

[0039] Based on credibility rating Classify scheduling reliability levels. If If so, only simulated scheduling suggestions are generated, and on-chain scheduling certificates are not issued; Then, candidate solutions are generated and entered into manual review; if If the event is deemed credible, then the smart contract verification and credential issuance process is allowed. Therefore, the trust level not only indicates the credibility of the event but also directly controls whether subsequent processes allow the automatic issuance of scheduling credentials.

[0040] Credible rating It is not only used for hierarchical gating, but also passed as a dynamic parameter to smart contracts to drive the dynamic adjustment of contract verification rules. This linkage mechanism enables the system to automatically tighten execution conditions under low-confidence events and simplify the process under high-confidence events, achieving risk-adaptive trusted execution.

[0041] Filter the set of facilities associated with the event: Based on the location of the event and the hydrological connectivity, the pollution impact range is calculated in the watershed management target map. For each candidate propagation path Calculate the expected pollution impact intensity of the target river section or cross-section: , in, The initial abnormal intensity of the pollution source, The attenuation coefficient is... This is the propagation delay. If... If the impact exceeds the preset threshold, then the node will be... Included in the scope of pollution impact.

[0042] For each treatment facility Read its facility's effective range For example, the effective range of a gate can be represented by the affected river section after the gate is closed, the length of the backwater, and the boundary of water level changes; the effective range of a pumping station can be represented by the affected river section, the flow regulation capacity, and the response period; the effective range of a sewage treatment facility can be represented by the pollution load it can accept, the remaining treatment capacity, and the service area; and the effective range of a water intake can be represented by the water intake protection zone and the area affected by water outages.

[0043] Facilities will be provided when the following conditions are met. Add to event-related facility collection : , in, This indicates the permission determination result obtained by querying based on scheduling permissions. Indicates the facility at any time The availability status. The availability status is determined based on at least the operating status, fault status, maintenance status, manual lockout status, communication status, and security boundaries.

[0044] This step transforms the map data into a specific set of event-related facilities, and the screening results are directly used to construct capability profiles. Compared to traditional manual screening, this step significantly improves the objectivity and accuracy of the screening through quantitative calculation of the scope of impact and triple determination of permissions / availability status.

[0045] Constructing a profile of governance facility scheduling capabilities: Associate facilities for each event Building a profile of scheduling capabilities:

[0046] in, For facility type; A set of executable actions; For action parameter boundaries; For response time; For the cost of the action; For ecological constraints; For safety constraints; Historical execution credit score; As the responsible party; This serves as a source of facility status data.

[0047] The executable actions can be enumerated according to the facility type. For example, the actions of a gate include opening, closing, and adjusting the opening degree; the actions of a pumping station include starting, stopping, and frequency adjustment; the actions of a sewage treatment facility include increasing the treatment load, switching to emergency processes, and accepting temporary pollution loads; the actions of a storage facility include storing water, releasing water, diverting floodwater, or regulating water storage; and the actions of a water intake include reducing water intake, suspending water intake, and switching to a backup water source.

[0048] The action parameter boundary can be represented as: , in, and These represent the upper and lower limits of the action parameters, respectively. Indicates the adjustment step size. This indicates the maximum rate of change.

[0049] The facility dispatch capability profile can be expressed using the following JSON structure: { "facility_id": "gate_01", "facility_type": "gate", "actions": ["open", "close", "adjust_opening"], "parameter_boundary": {"opening_min": 0.0, "opening_max": 2.5, "step": 0.1}, "response_time": {"start_delay_min": 5, "finish_delay_min": 20}, "safety_constraints": ["upstream_level<= warning_level"], "ecological_constraints": ["downstream_flow>= ecological_flow_min"], "credit": 0.86, "owner": "basin_dispatch_center", "state_oracle": "trusted_gateway_01" }

[0050] This step enriches the filtered facilities from equipment numbers into computable scheduling entities. (Image) The initial value can be obtained from historical records or set as a default value, and then continuously updated before being input back into this step. The profile also serves as the basis for bidding capability in contract network negotiation and the simulation boundary input for counterfactual assessment, making it a pivotal node in the entire closed-loop system.

[0051] like Figure 3 As shown, candidate governance scheduling schemes are generated based on the contract network negotiation mechanism: The scheduling task corresponding to a sudden water pollution event is represented as follows: , in, It should include at least pollution peak reduction targets, sensitive target protection targets, and response time targets; It should include at least scheduling authority constraints, facility action boundaries, safety constraints, and ecological constraints.

[0052] Event scheduling agents move towards event-related facility sets The task announcement is issued. Each facility agent generates bidding information based on its own scheduling capability profile: , in, The proposed action and parameters; Contribution to expected governance; This is the estimated response time; For the cost of the action; For safety or ecological risks; Historical execution credit score; Signatures or commitments by the responsible party for the facility.

[0053] Calculate a negotiated score for each bid: , in, to The historical execution credit score is used for weighting. It is updated not only after the fact but also during the negotiation phase, influencing bid ranking and proposal weighting. Facilities with credit scores below a preset freeze threshold are not automatically scheduled but are considered for manual review.

[0054] The candidate solution consists of several facility actions: , in, The system prioritizes facility actions within the plan. It generates either the top-K candidate plans or a set of Pareto candidate plans that satisfy the constraints.

[0055] When there are conflicts in the actions of candidate facilities, a conflict constraint matrix is ​​used. Judgment. For example, upstream sluice gate closure leading to insufficient downstream ecological flow, conflicts between pumping station discharge and intake protection targets, and insufficient capacity of sewage treatment facilities. If Then you cannot select facilities at the same time. and The system will arbitrate the corresponding actions based on the risk of sensitive targets, scheduling trust level, credit value and expected governance contribution; if arbitration is not possible, renegotiation will be triggered.

[0056] This step enables the automatic generation of multi-facility collaborative solutions. Compared with classic contract agreements, the improvements of this application are: the bid scoring incorporates historical credit scores; renegotiation is triggered after conflict arbitration rather than a simple rejection; and the results of counterfactual assessment can be used to correct the weight and historical credit scores of this step.

[0057] Conduct counterfactual effect evaluation on candidate solutions: For candidate solutions Constructing counterfactual scenarios The effectiveness of the proposed solutions is evaluated using hydrodynamic and water quality models, pollution diffusion proxy models, or historical similarity event proxy models. The counterfactual effect score can be expressed as: , in, Indicates the peak pollution reduction rate. This indicates the delay time for contamination to reach sensitive targets. Indicates the cost of facility operation. Indicates safety or ecological risk. Indicates the degree of constraint violation.

[0058] The peak pollution reduction rate can be expressed as: , in, The peak pollution level when the candidate scheme is not implemented. The predicted pollution peak after implementing the candidate scheme.

[0059] like If the candidate solution falls below a preset threshold, or violates action boundaries, safety constraints, or ecological constraints, the counterfactual evaluation module rejects the solution and returns the reason for rejection. If there are locally correctable parameters, the candidate action is adjusted according to the action parameter boundaries and re-evaluated. If there are no feasible corrections, the process returns to the previous step for renegotiation.

[0060] The evaluation results directly influence the negotiation parameters through feedback instructions, enabling the multi-agent system to "learn" from simulations and progressively improve the integrity of bids and the feasibility of solutions. This is fundamentally different from the use of simulations in existing technologies, where "simulation is only used for post-event verification" or "only as a gating mechanism."

[0061] like Figure 4 As shown, on-chain scheduling credentials are verified and generated through a blockchain smart contract: Candidate solutions that pass the counterfactual evaluation will be submitted to the blockchain smart contract. The smart contract will perform at least the following checks: Does the entity initiating the dispatch have dispatch authority over the event area and facility set? Do the candidate proposals include the signatures or collaborative commitments of the relevant responsible parties for the facilities? Whether the available status reported by the facility status oracle is consistent with the candidate solution; Are the facility's motion parameters within the motion parameter boundaries? Do the candidate schemes meet the rules regarding ecological flow, water level safety, water intake protection, and cross-regional notification? Does the counterfactual assessment result hash match the off-chain assessment report?

[0062] Cross-regional collaboration rules can be formalized as follows: , in, For the set of cross-regional entities affected by the candidate solutions, Representing the subject Dispatch notification received Representing the subject A collaborative commitment has been signed.

[0063] After verification, the smart contract generates an on-chain scheduling credential: { "credential_id": "cred_20260429_001", "event_id": "event_001", "plan_hash": "hash(plan_m)", "facility_actions_hash": "hash(actions)", "expected_effect_hash": "hash(counterfactual_report)", "participants": ["dispatch_center", "gate_01_owner", "plant_02_owner"], "signatures": ["sig_dispatch", "sig_gate", "sig_plant"], "valid_time": ["2026-04-29T10:00:00", "2026-04-29T18:00:00"], "status": "issued" } This on-chain scheduling certificate is not merely data storage; rather, it binds multi-party collaborative commitments, scheduling permissions, facility action boundaries, and expected governance effects into an undeniable basis for execution. If verification fails, the smart contract returns a failure code and a reason for failure, and the system rolls back based on the reason.

[0064] This step achieves the final implementation of trusted execution. Compared to existing blockchain-based evidence storage systems, the improvement of this application lies in: the certificate content includes a complete scheduling scheme, commitment, and expected results, rather than a simple data hash; the contract rules are subject to... Dynamic modulation enables risk self-adaptation; facility actions in the voucher are bound to the profile to ensure traceability of execution boundaries.

[0065] like Figure 5 As shown, the process includes verifying receipts, calculating deviations in governance effectiveness, and updating credit scores. After executing a scheduling instruction, the facility uploads an execution receipt. The execution receipt must include at least the facility identifier, action type, actual action parameters, start time, completion time, executor signature, on-site data hash, and an exception description. The system verifies that the execution receipt matches the on-chain scheduling credential. , in, This indicates that the actual action is consistent with the action on the voucher. This indicates that the execution time satisfies the voucher constraint. This indicates that the signature is valid.

[0066] The deviation in governance effectiveness was calculated based on multi-source feedback data.

[0067] in, To predict the pollution reduction rate, To measure the actual pollution reduction rate, To predict response time, This is the actual response time. For the degree of actual violation of constraints, For the sake of completeness of the receipt.

[0068] Update the historical performance credit scores of participating facilities: , in, To preserve the historical coefficient, Assess the weight of the facility's responsibility or contribution in the plan. Penalties for failure to execute according to the voucher, execution exceeding the time limit, or false receipts. This means that the credit score will be limited to between 0 and 1.

[0069] Updated The credit score is entered into the facility's scheduling capability profile, influencing subsequent negotiation scores, facility freeze determination, and manual review requirements. When the credit score is below the freeze threshold, the system marks the facility as having a restricted scheduling status; when the credit score is above the reward threshold, the facility receives a higher proposal weight in subsequent negotiations.

[0070] The technical solution of this application is not a simple, sequential overlay of knowledge graphs, weighted scoring, contract network negotiation, counterfactual simulation, and blockchain smart contracts. Instead, it addresses the unique challenges of "cross-regional, multi-stakeholder, dynamic constraints, and trusted collaboration" in sudden water pollution events in river basins by constructing a closed-loop system with deep coupling and mutual feedback among multiple technical features. The synergistic effect among these features produces a technical effect that exceeds the sum of their functions, specifically manifested in the following key coupling points: Dynamic linkage between scheduling trust level and smart contract verification: Scheduling Trust Level It not only serves as a gating threshold for whether to enter automatic issuance, but also as a dynamic parameter directly written into the smart contract's verification logic. Specifically: when At a moderate level (e.g.) When this happens, the smart contract automatically adds additional verification conditions, such as requiring offline signature confirmation from at least two cross-regional entities, or requiring a 10% increase in the confidence level of the counterfactual assessment; when When the level is high, the verification process is simplified. This "trust level-driven adaptive adjustment of smart contract rules" mechanism enables the system to ensure prudent control under high-risk events and achieve rapid response under high-trust events. In existing technologies, trust levels are only used for display or simple gating and have never been dynamically bound to the smart contract's rule engine, thus producing the unexpected technical effect of "risk-adaptive trusted execution".

[0071] The dual constraints of facility dispatch capacity profiling and counterfactual assessment: Historical execution credit value in facility dispatch capability profile and safety / ecological constraints These are not static parameters, but rather dynamic boundary conditions and execution noise factors for the counterfactual simulation model. Specifically: in counterfactual evaluation, for credit scores... For facilities below a preset threshold, the model will automatically apply an additional random delay perturbation to their actions (the perturbation amplitude is...). (proportional to) and reduce the weight of its expected contribution value; at the same time, safety constraints The credit score and security constraints are converted into hard constraints in the simulation model, and the evaluation of the scheme terminates once they are violated. This mechanism of dynamically injecting credit scores and security constraints into counterfactual simulations enables the evaluation results to proactively reflect the facility's historical unreliability and real-time risks, solving the deficiency of traditional simulation models that cannot integrate facility credibility information. Existing counterfactual assessments or simulations are based on idealized facility parameters and cannot handle execution deviations caused by differences in the creditworthiness of the entities. This application fills this technological gap through this coupling mechanism.

[0072] The counterfactual assessment results directly corrected the contract negotiation process: When a counterfactual assessment determines that a candidate solution does not meet constraints or has insufficient effectiveness, the system not only rejects it but also encodes specific problems identified in the assessment (such as an inflated contribution of a facility or a conflicting combination of actions) as correction instructions and feeds them back to the contract network for negotiation. For example, if a pumping station's bid commitment reduction rate is found to be too high in the counterfactual assessment (unable to be achieved due to downstream water level backflow), the system automatically lowers the pumping station's credit score weight in the next round of negotiation. Furthermore, it mandates that bidders provide a more conservative contribution estimate when resubmitting their bids, while also adding a virtual penalty cost. This closed-loop feedback mechanism, where evaluation results correct negotiation parameters, allows multi-agent negotiation to learn from simulation experience and optimize bid integrity and solution feasibility round by round. In existing contract network protocols, bid evaluation is based solely on the bid value itself, lacking a feedback correction mechanism based on global simulation. The design in this application significantly improves the probability of negotiation converging to a truly feasible solution.

[0073] The closed-loop effect of historical execution credit scores: from post-event evaluation to pre-event screening and access control.

[0074] Updated historical execution credit value This not only affects subsequent negotiation scoring, but also directly drives two types of pre-control measures: first, a facility freeze threshold, where facilities below this threshold are automatically downgraded from the "automatic scheduling facility set" to "candidate facilities for manual review only," or even temporarily removed from the associated facility set. Secondly, scheduling permissions are dynamically adjusted. For facilities with excessively low credit scores, the scheduling permissions of the responsible entity will be temporarily set to "read-only" by the smart contract, preventing them from signing new scheduling commitments. This credit-driven adaptive contraction mechanism for facility autonomy permissions enables the system to automatically isolate unreliable executors without manual intervention, significantly improving system robustness in multi-round, cross-event scheduling. Existing blockchain-based evidence storage systems only record credit and do not use credit scores as input for real-time control of scheduling qualifications and permissions. This application achieves real-time closed-loop control of the scheduling process through credit evaluation.

[0075] In summary, the various technical features of this application exhibit a bidirectional, nonlinear, and dynamic feedback synergistic relationship, rather than a simple, unidirectional, sequential superposition. Specifically: Capability profiling (including credit score) → Constraint and disturbance counterfactual assessment → Modified parameter contract network negotiation → Generation of solution smart contract → Execution feedback to update credit score → Re-constrain capability profiling. This closed-loop chain, in the cross-regional, multi-entity, and dynamically constrained scenario of sudden water pollution events in a watershed, achieves for the first time the organic unity of "trustworthy classification - dynamic modeling - collaborative negotiation - forward-looking assessment - trustworthy execution - credit feedback," producing technical effects such as "adaptive risk control," "self-learning negotiation optimization," and "credit-driven permission downgrading" that surpass the sum of the individual functions of each module.

[0076] This application also includes a credit-driven trusted collaborative scheduling system for sudden water pollution events in a watershed, and implements a credit-driven trusted collaborative scheduling method for sudden water pollution events in a watershed, including: The watershed governance object map construction module is used to construct a watershed governance object map that includes river sections, monitoring sections, sewage outlets, gates, pumping stations, sewage treatment facilities, storage facilities, water intakes, administrative regions, responsible entities, and relationships such as hydrological connectivity, pollution propagation, facility functions, and scheduling authority. The scheduling credibility level calculation module is used to calculate the scheduling credibility score based on the abnormal intensity of pollutants, the duration of abnormality, the consistency of hydrological propagation, the risk of sensitive targets, data integrity, and the status of manual verification. The event-related facility filtering module is used to filter the set of event-related facilities based on the scope of pollution impact, the scope of facility operation, scheduling authority relationships, and facility availability status. The facility scheduling capability profile building module is used to build a scheduling capability profile for each facility in the event-related facility set, including a set of executable actions, action parameter boundaries, response time, security constraints, ecological constraints, and historical execution credit value. The contract negotiation module is used to generate candidate governance scheduling schemes based on scheduling capability profiles and historical execution credit values, through facility intelligent agents bidding and negotiation scoring ranking. The counterfactual assessment module is used to predict the effects and verify the constraints of candidate governance scheduling schemes, and to reject or modify candidate schemes or trigger the contract network negotiation module to renegotiate based on the assessment results. The blockchain smart contract module is used to verify the scheduling authority, responsible entity signature, facility status, action parameter boundaries, and cross-regional collaborative commitment of candidate solutions. After the verification is passed, an on-chain scheduling certificate is generated. The on-chain scheduling certificate includes at least the solution hash, multi-party signatures, validity period, and execution conditions. The execution verification and credit update module is used to obtain execution receipts and multi-source feedback data, verify the consistency between execution receipts and on-chain scheduling credentials, calculate governance effect deviation and update historical execution credit values ​​based on governance effect deviation, and write the updated historical execution credit values ​​back to the facility scheduling capability profile building module. The system is configured as follows: when the updated historical execution credit value is lower than the preset freezing threshold, the event-related facility screening module excludes the facility from the set of automatically scheduled facilities; the scheduling trust score is input as a dynamic parameter into the blockchain smart contract module to dynamically adjust the stringency of the contract verification rules.

[0077] The invention will be further explained below with reference to a sudden water pollution incident in the river basin. This embodiment is used to illustrate the technical solution of the invention and does not constitute a limitation on the scope of protection.

[0078] In one embodiment, the trusted collaborative governance scheduling of sudden pollution events includes: An abnormally high ammonia nitrogen level occurred in a watershed. Upstream of the event were sewage outlets, sluice gates, pumping stations, and wastewater treatment facilities, while downstream were water intakes and sensitive cross-sections. The system first located the event node in the watershed management object map based on monitoring cross-sections and hydrological data, and then calculated the pollution propagation path and estimated arrival time along the hydrological connectivity edges.

[0079] The system calculates the anomaly intensity, anomaly duration, hydrological propagation consistency, risk to sensitive targets, data integrity, and manual verification status to obtain a scheduling reliability score. If the score reaches the automatic scheduling threshold, the facility screening process begins; if the score is within the manual review range, candidate suggestions are generated but on-chain scheduling certificates are not issued; if the score is below the simulation threshold, only simulation is performed.

[0080] The system screens candidate facilities based on the scope of pollution impact and the facility's operational range. For gates, the system reads their adjustable opening range, maximum adjustment rate, upstream water level safety constraints, and downstream ecological flow constraints; for pumping stations, the system reads their maximum pumping capacity, start-stop delay, and power status; for wastewater treatment facilities, the system reads their remaining treatment capacity, emergency process status, and acceptance boundaries; and for water intakes, the system reads their reduction or shutdown boundaries and backup water source status.

[0081] The event scheduling agent publishes contract network task announcements to candidate facility agents. Each facility agent submits a bid based on its own capability profile, including the proposed action, expected pollution reduction contribution, response time, action cost, risk indicators, credit score, and the signature of the responsible party. The system ranks the bids according to a negotiation scoring function and generates several candidate scheduling schemes based on the conflict constraint matrix.

[0082] For candidate solutions, the system constructs counterfactual scenarios to predict downstream pollution peaks, arrival times at sensitive targets, changes in ecological flow, and facility operation risks after implementing the solution. If a candidate solution can reduce pollution peaks and meet safety and ecological constraints, it is submitted for smart contract verification; if a candidate solution does not meet the constraints, the counterfactual assessment module provides reasons for rejection or parameter correction suggestions and triggers renegotiation.

[0083] After receiving candidate solutions, the smart contract verifies the scheduling initiator's permissions, the facility's responsible entity's signature, the facility's state oracle data, action parameter boundaries, and cross-regional coordination rules. Upon successful verification, an on-chain scheduling credential is generated. Each facility executes the action according to the credential and uploads a signed execution receipt.

[0084] After execution, the system calculates the deviation in treatment effectiveness based on water quality monitoring data, hydrological feedback data, facility status data, and manual verification data. If the actual reduction effect is close to the predicted effect and the execution receipt is complete, the credit value of the relevant facility is increased; if the execution timeout, action parameters are inconsistent, the effect deviates significantly, or the receipt is missing, the credit value of the relevant facility is decreased. The updated credit value is written back to the facility's dispatch capability profile and affects bidding ranking and dispatch eligibility in subsequent contract network negotiations.

[0085] In another embodiment, the rollback process after a smart contract verification failure includes: If the smart contract verification finds that a pump station is in a maintenance lock state, it returns a facility status failure code, removes the pump station from the event-associated facility set, and re-filters the facilities.

[0086] If the smart contract verification finds that a gate action exceeds the opening boundary, it returns an action boundary failure code, and the system corrects the action parameters of the candidate solutions.

[0087] If the smart contract verification finds that the downstream water intake entity has not signed a cross-regional collaborative commitment, a collaborative commitment failure code will be returned, and the system will re-initiate contract network negotiation or enter manual review.

[0088] Through the aforementioned fallback path, the system can explain the specific triggering conditions and processing procedures for the rejection, modification, or renegotiation of candidate solutions.

[0089] In another specific embodiment, the implementation scenario, data interface, and parameter calibration allocation are as follows: like Figure 6 As shown, in one implementation scenario, upstream sewage outlets, upstream monitoring sections, sluice gates, pumping stations, sewage treatment facilities, water storage facilities, downstream sensitive sections, and water intakes are respectively represented as nodes in the watershed governance object map. Hydrological connectivity edges are used to represent pollution propagation paths, facility function edges are used to represent the scope of each facility's effect on pollution reduction, pollution delay, flow regulation, or water intake protection, and scheduling authority edges are used to represent the authority of the regulatory or responsible entity to initiate, confirm, and sign off on facility actions.

[0090] In this embodiment, the multi-source data interfaces include a water quality monitoring interface, a hydrological station interface, an online discharge outlet monitoring interface, a SCADA facility status interface, an equipment ledger interface, a scheduling procedure interface, an ecological flow control rule interface, and a manual verification interface. The water quality monitoring interface provides pollutant concentration, sampling time, and cross-sectional location; the hydrological station interface provides flow velocity, flow rate, water level, and propagation delay correction parameters; the SCADA facility status interface provides gate opening, pump station start / stop status, remaining treatment capacity of wastewater treatment facilities, remaining storage capacity of regulating facilities, and communication status; the equipment ledger and scheduling procedures provide action boundaries, maximum rate of change, response time, safe water level, and ecological flow constraints.

[0091] The credible scoring weights can be initially determined using historical samples of sudden water pollution events and expert rules, and then adjusted based on the event review results. Threshold and The system determines the weights and thresholds based on the false positive rate, false negative rate, manual review capacity, and risk level of automatic issuance. If the historical sample size is insufficient, the system records the conclusions of each manual review, the on-chain certificate issuance results, and the deviation in execution effect, and recalibrates the weights and thresholds after the preset sample size is reached.

[0092] The facility capability profile records the data source and timestamp simultaneously during generation. For automatically acquired data, the profile fields are bound to the corresponding state oracle identifier; for action boundaries or collaborative commitments that cannot be automatically acquired, the system requires the responsible party to confirm and generate a signature through a manual review interface. Thus, the capability profile serves not only as input for contract network negotiation but also as a basis for smart contracts to verify facility status and action boundaries.

[0093] In another embodiment, pre-issuance gatening, failure fallback, and credit feedback include: like Figure 7 As shown, the system first calculates the scheduling reliability score after an event input. .when Less than the first threshold At that time, the system only generates simulated scheduling suggestions and items to be verified; when Located at the first threshold Second threshold In the meantime, the system allows the generation of candidate solutions, but these must be manually reviewed; when Not less than the second threshold At that time, the system allows candidate solutions to enter the counterfactual evaluation and smart contract verification process.

[0094] The counterfactual assessment module receives candidate governance scheduling schemes, pollution source strength, flow rate, water level, propagation delay, facility action parameters, sensitive target location, action cost, safety constraints, and ecological constraints as inputs. It outputs the pollution peak reduction rate, the delay time to reach the sensitive target, cost, risk, and the degree of constraint violation. When the counterfactual score is below a preset threshold, the system returns the candidate scheme number, the reason for failure, and correctable parameters; if only local parameters exceed limits, it returns action correction; if incompatible facility actions exist, it returns for contract network renegotiation.

[0095] When smart contract verification fails, the system executes a rollback according to the failure code: if the facility status is unavailable, it returns to the facility screening step; if the action parameters are out of bounds, it returns to the counterfactual assessment step for parameter correction; if there is a lack of cross-regional collaborative commitment or multi-party signatures, it returns to the contract network negotiation step; if the counterfactual assessment hash is inconsistent, it refuses to issue on-chain scheduling credentials and triggers manual review. The above rollback path ensures that every rejection, correction, or renegotiation has a traceable triggering condition.

[0096] After the facility performs its work, it uploads an execution receipt. The system verifies the actual action parameters, start time, completion time, execution entity signature, and on-site data hash. If the receipt is missing, the signature is invalid, the action parameters are inconsistent with the voucher, the execution times out, or the measured treatment effect is significantly lower than the predicted effect, the system lowers the corresponding facility's credit value and marks the facility as a candidate for manual review or in a restricted scheduling state. If the receipt is complete and the deviation between the predicted and measured effects is less than a preset threshold, the system raises the corresponding facility's credit value, giving it a higher bidding weight in subsequent negotiations.

[0097] Quantitative Record of Technical Effects: To support the technical effectiveness, this embodiment records at least four types of quantitative indicators after each event handling: the number and rate of unauthorized scheduling interceptions, the number and rate of candidate scheme constraint violations, the deviation between the predicted pollution reduction rate and the measured pollution reduction rate, the execution receipt integrity rate, and the signature validity rate. These indicators correspond one-to-one with the trust level gating, counterfactual pre-issuance gating, smart contract verification, and credit feedback steps, demonstrating that this invention reduces the probability of unauthorized scheduling, non-executable scheduling, and insufficiently effective scheduling entering the execution phase.

[0098] The embodiments described above merely illustrate specific implementation methods of this application, and while the descriptions are detailed and specific, they should not be construed as limiting the scope of protection of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the technical solution of this application, and these modifications and improvements all fall within the scope of protection of this application.

Claims

1. A credit-driven, reliable collaborative scheduling method for sudden water pollution events in a watershed, characterized in that, Includes the following steps: Construct a map of watershed governance objects, which includes river section nodes, monitoring section nodes, sewage outlet nodes, gate nodes, pumping station nodes, sewage treatment facility nodes, storage facility nodes, water intake nodes, administrative region nodes, responsible entity nodes, as well as hydrological connectivity edges, pollution propagation edges, facility function edges, dispatch authority edges, local management edges, and data reporting edges. Acquire multi-source sensing data on sudden water pollution events, and calculate scheduling reliability scores based on pollutant abnormal intensity, abnormal duration, consistency of hydrological propagation, risk of sensitive targets, data integrity, and manual verification status. The pollution impact range is calculated based on the location of the event and the hydrological connectivity edges in the watershed management object map. The set of facilities associated with the event is then selected based on whether there is an intersection between the pollution impact range and the facility's operational range, the scheduling authority determination results, and the facility's availability status. For each facility in the event-related facility set, construct a scheduling capability profile. The scheduling capability profile shall include at least the set of executable actions, action parameter boundaries, response time, security constraints, ecological constraints, and historical execution credit value. Based on scheduling capability profiles and historical execution credit scores, a contract network negotiation mechanism is used to generate candidate governance scheduling schemes. Conduct counterfactual effect assessments on candidate governance scheduling schemes, and reject, revise, or trigger renegotiation based on the assessment results; Candidate solutions that pass the counterfactual effect evaluation are submitted to the blockchain smart contract for verification. Once the verification is passed, an on-chain scheduling certificate is generated. Obtain execution receipts and multi-source feedback data after facility execution, verify the consistency between execution receipts and on-chain scheduling credentials, calculate governance effect deviation, update historical execution credit values ​​based on governance effect deviation, and write the updated historical execution credit values ​​back to the scheduling capability profile.

2. The credit-driven trusted collaborative scheduling method for sudden water pollution events in a watershed, as described in claim 1, is characterized in that... The process of updating historical execution credit values ​​based on governance performance deviations and writing the updated historical execution credit values ​​back to the scheduling capability profile also includes: The updated historical execution credit value has a reverse impact on the bidding score weight of the corresponding facility in subsequent contract network negotiations; when the updated historical execution credit value is lower than the preset freezing threshold, the corresponding facility will be removed from the set of automatically scheduled facilities and marked as a candidate facility for manual review only; when the updated historical execution credit value is higher than the preset reward threshold, the bidding score weight of the corresponding facility will be increased in subsequent negotiations. The scheduling trust score is input into the blockchain smart contract as a dynamic parameter. When the scheduling trust score is in the preset medium range, the smart contract automatically adds additional cross-regional subject signature verification conditions.

3. The credit-driven trusted collaborative scheduling method for sudden water pollution events in a watershed, as described in claim 1, is characterized in that... The counterfactual effect evaluation of candidate governance scheduling schemes, and the rejection, modification, or triggering of renegotiation based on the evaluation results, specifically includes: Construct counterfactual scenarios corresponding to candidate governance scheduling schemes, and predict the peak pollution reduction rate and the arrival delay time of sensitive target pollution after implementing the scheme through hydrodynamic water quality models or pollution diffusion proxy models. When the counterfactual effect score is lower than the preset threshold or the scheme violates safety or ecological constraints, if there are correctable parameters, the facility action parameters in the candidate scheme are adjusted according to the action parameter boundary and re-evaluated; if there are no feasible corrections, the specific problems identified in the evaluation are encoded as correction instructions and returned to the contract network negotiation mechanism, the credit score weight of the corresponding facility in the next round of negotiation is reduced, and it is forced to provide a corrected contribution estimate when re-bidding.

4. The credit-driven trusted collaborative scheduling method for sudden water pollution events in a watershed, as described in claim 1, is characterized in that... The calculation of the scheduling reliability score based on pollutant anomaly intensity, anomaly duration, hydrological propagation consistency, sensitive target risk, data integrity, and manual verification status specifically includes: Calculate the scheduling reliability score using the following formula: ， The abnormal intensity of pollutants is calculated as follows: ， in, For actual measured concentration, Background concentration, These are the standard limits for the corresponding pollutants; The duration of the anomaly is calculated as follows: ， in, The duration of continuous anomalies, The preset duration threshold; Hydrological propagation consistency is calculated as follows: ， in, The time when the anomaly was observed at the downstream monitoring point. The arrival time is predicted based on upstream anomalies and propagation delays. This is the allowable deviation threshold; When the scheduling trust score is lower than the first threshold, only a simulated scheduling suggestion is generated and no on-chain scheduling certificate is issued; when the scheduling trust score is between the first threshold and the second threshold, a candidate scheme is generated and enters the manual review process; when the scheduling trust score is not lower than the second threshold, the process of smart contract verification and certificate issuance is allowed.

5. The credit-driven trusted collaborative scheduling method for sudden water pollution events in a watershed, as described in claim 1, is characterized in that... The process of verifying the consistency between the execution receipt and the on-chain scheduling certificate, calculating the governance effect deviation, and updating the historical execution credit value based on the governance effect deviation includes: Verify the consistency of operation of each facility as follows: ， in, This indicates that the actual action is consistent with the action on the voucher. This indicates that the execution time satisfies the voucher constraint. This indicates that the signature of the executing entity is valid; Calculate the treatment effect deviation using the following formula: in, To predict the pollution reduction rate, To measure the actual pollution reduction rate, To predict response time, This is the actual response time. For the degree of actual violation of constraints, For the sake of completeness of the receipt; Update historical execution credit scores using the following formula: ， in, To preserve the historical coefficient, Assess the responsibility weight of the facility in the plan. Penalties for failure to execute according to the voucher, execution exceeding the time limit, or false receipts. This means that the credit score will be limited to between 0 and 1.

6. The credit-driven trusted collaborative scheduling method for sudden water pollution events in a watershed, as described in claim 1, is characterized in that... The process of generating candidate governance scheduling schemes based on scheduling capability profiles and historical execution credit values ​​using a contract network negotiation mechanism specifically includes: The event scheduling agent publishes task announcements to the event-related facility set. Each facility agent generates bidding information based on its own scheduling capability profile. The bidding information includes the proposed action and parameters, expected governance contribution, expected response time, action cost, security or ecological risks, and historical execution credit value. The negotiation score for each bid is calculated using the following formula: ， in, To the expected contribution to governance, To estimate the response time, As a cost of action, For safety or ecological risks, For historical execution credit scores, to As weight; The bids are ranked based on the negotiated scores, generating the top-K candidate solutions or a set of Pareto candidate solutions that meet the constraints.

7. A credit-driven trusted collaborative scheduling method for sudden water pollution events in a watershed, as described in claim 6, is characterized in that... After generating the candidate governance scheduling scheme, the following is also included: A conflict constraint matrix is ​​used to determine conflicts between candidate facility actions. Indication facilities and facilities The corresponding actions cannot be selected simultaneously; when a conflict is determined, a weighted arbitration is conducted based on the risk of sensitive targets, scheduling credibility score, historical execution credit value and expected governance contribution. If the conflict cannot be resolved after arbitration, renegotiation is triggered, requiring both parties to submit revised bids.

8. The credit-driven trusted collaborative scheduling method for sudden water pollution events in a watershed, as described in claim 1, is characterized in that... The initial value of the historical execution credit value in the scheduling capability profile is determined by the statistics of historical scheduling records or set as the default value. The action parameter boundaries include the minimum value, maximum value, adjustment step size and maximum rate of change of the action parameters. The safety constraints include the upstream water level warning constraints and the ecological constraints include the downstream ecological flow lower limit constraints.

9. A credit-driven trusted collaborative scheduling method for sudden water pollution events in a watershed, as described in claim 1, is characterized in that... The process of filtering the set of facilities associated with the event, based on whether there is an overlap between the pollution impact range and the operational range of each treatment facility, the scheduling authority determination result, and the facility availability status, specifically includes: The facility will be determined according to the following criteria. Add to event-related facility collection : ， in, The extent of pollution impact, For the area of ​​effect of the facility, This refers to the permission determination result obtained by querying based on scheduling permissions. For facilities at all times The availability status; The availability status of a facility is determined by a combination of factors, including its operational status, fault status, maintenance status, manual lockout status, communication status, and security boundaries. If any of these statuses indicate that a facility is unavailable, it is excluded from the set of facilities associated with the event.

10. A credit-driven, reliable collaborative scheduling system for sudden water pollution events in a watershed, characterized in that, Implementing a credit-driven trusted collaborative scheduling method for sudden water pollution events in a watershed, as described in any one of claims 1-9, includes: The watershed governance object map construction module is used to construct a watershed governance object map that includes river sections, monitoring sections, sewage outlets, gates, pumping stations, sewage treatment facilities, storage facilities, water intakes, administrative regions, responsible entities, and relationships such as hydrological connectivity, pollution propagation, facility functions, and scheduling authority. The scheduling credibility level calculation module is used to calculate the scheduling credibility score based on the abnormal intensity of pollutants, the duration of abnormality, the consistency of hydrological propagation, the risk of sensitive targets, data integrity, and the status of manual verification. The event-related facility filtering module is used to filter the set of event-related facilities based on the scope of pollution impact, the scope of facility operation, scheduling authority relationships, and facility availability status. The facility scheduling capability profile building module is used to build a scheduling capability profile for each facility in the event-related facility set, including a set of executable actions, action parameter boundaries, response time, security constraints, ecological constraints, and historical execution credit value. The contract negotiation module is used to generate candidate governance scheduling schemes based on scheduling capability profiles and historical execution credit values, through facility intelligent agents bidding and negotiation scoring ranking. The counterfactual assessment module is used to predict the effects and verify the constraints of candidate governance scheduling schemes, and to reject or modify candidate schemes or trigger the contract network negotiation module to renegotiate based on the assessment results. The blockchain smart contract module is used to verify the scheduling authority, responsible entity signature, facility status, action parameter boundaries, and cross-regional collaborative commitment of candidate solutions. After the verification is passed, an on-chain scheduling certificate is generated. The on-chain scheduling certificate includes at least the solution hash, multi-party signatures, validity period, and execution conditions. The execution verification and credit update module is used to obtain execution receipts and multi-source feedback data, verify the consistency between execution receipts and on-chain scheduling credentials, calculate governance effect deviation and update historical execution credit values ​​based on governance effect deviation, and write the updated historical execution credit values ​​back to the facility scheduling capability profile building module. The system is configured as follows: when the updated historical execution credit value is lower than the preset freezing threshold, the event-related facility screening module excludes the facility from the set of automatically scheduled facilities; the scheduling trust score is input as a dynamic parameter into the blockchain smart contract module to dynamically adjust the stringency of the contract verification rules.

Images