Renewable resource recovery system and method based on big data

Abstract

The invention relates to the field of resource recovery processing, and discloses a renewable resource recovery system and method based on big data, and the system comprises a data collection module, a resource recovery analysis module, and a planning and scheduling module. Through multi-source data acquisition, a structured data set is formed after preprocessing fusion; energy cost is accurately calculated based on a vehicle dynamics model, service urgency degree cost is quantified through a regional gridding model, and time cost is estimated in combination with real-time traffic; predicting the global state of the system by analyzing the time sequence slope sequence features; according to a system state, utilizing an adaptive weight model to dynamically generate various cost weights, constructing a weighted comprehensive cost function, and utilizing a path optimization algorithm to solve an optimal collection and transportation path sequence under the condition that load and task constraint conditions are met; and finally decomposing the path sequence into operation instructions and issuing the operation instructions to the vehicle terminal. According to the invention, the crossing from static planning to dynamic adaptive scheduling is realized, and the economical efficiency, efficiency and service quality of resource recovery operation are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of resource recycling, and in particular to a renewable resource recycling system and method based on big data. BACKGROUND

[0002] Efficient recycling and utilization of renewable resources is a key link to promote the development of circular economy and realize the sustainable operation of cities. It not only concerns resource conservation and environmental protection, but also has strategic significance for waste reduction and harmless treatment in cities. With the improvement of social consumption level, the amount of renewable resources is increasing, and higher requirements are put forward for the efficiency and intelligent level of the recycling system; At present, the recycling operation of renewable resources mainly depends on two modes: one is a planned collection mode with fixed routes and fixed cycles, and the recycling vehicles collect according to the preset routes; the other is to receive orders through artificial telephone or simple information platform, and the dispatch personnel arrange temporary orders and paths according to experience. These methods realize the preliminary collection of resources to a certain extent, but they are passive and extensive management methods; However, the above existing methods have obvious deficiencies and defects. The fixed route mode is rigid and cannot respond to the dynamic changes of inventory and sudden task requirements of the recycling point, which easily leads to high empty load rate of vehicles, and full load recycling points cannot be cleared in time. The dispatching method relying on artificial experience has slow decision-making process and strong subjectivity, and it is difficult to comprehensively optimize vehicle energy consumption, time cost and service response speed and other multiple targets. Especially in a large-scale recycling network, the existing technology lacks the ability of dynamic perception and quantitative analysis of real-time traffic, vehicle load, regional order backlog risk and other key factors, resulting in unscientific path planning, high operating cost, and bottleneck of overall recycling efficiency and service quality. SUMMARY

[0003] The purpose of the present application can be achieved by the following technical scheme: a renewable resource recycling system based on big data, comprising: A data acquisition module acquires static basic information and vehicle operation parameters of the recycling network through a system configuration database, Internet of Things sensing devices deployed at recycling nodes and vehicles, user terminal interfaces, and third-party map and energy service APIs. The original data collected are preprocessed to finally aggregate multi-source data into structured fusion data sets according to node number and time index. The resource recycling analysis module calculates the estimated total energy consumption based on a vehicle dynamics model and real-time parameters, and calculates the energy cost in combination with the unit price of energy; at the same time, the node urgency is calculated based on the grid region, historical data and order waiting time, and the region urgency is defined accordingly; the total task time is also estimated based on real-time traffic and node workload, and the final time cost is calculated in combination with the unit time cost coefficient; finally, the time sequence slope sequence fitted by analyzing the region urgency and order waiting time is extracted to obtain the feature vector, and the system state value and specific state are output by using the pre-set classifier; The planning and scheduling module quantifies the system state value into a specific value based on the received system state value, and inputs the specific value into an adaptive weight calculation model based on an exponential function and normalization processing together with a preset strategy bias coefficient to dynamically generate the weights of each cost; then, an optimization function with the minimum weighted comprehensive cost as the objective is constructed, and the optimal collection and transportation path sequence of the vehicle fleet is output by solving the optimization function based on a path optimization algorithm under the conditions of satisfying the vehicle load, task uniqueness and capacity constraints; finally, the path sequence is decomposed into navigation and operation instructions, which are sent to the vehicle terminal through a wireless communication network.

[0004] Preferably, the static basic information and vehicle operating parameters specifically include: The static basic information includes the unique identifier of all resource recycling nodes, GPS latitude and longitude coordinate information, and the theoretical maximum capacity value of the recycling nodes, the unique identifier of all resource recycling vehicles, the energy type of the vehicle, the baseline rolling resistance coefficient and air resistance coefficient of the vehicle; The vehicle operating parameters include the real-time vehicle load mass, rated load mass and empty vehicle weight uploaded by the resource recycling node, the equipment full load state flag, real-time position, driving speed, remaining load capacity, vehicle battery power or fuel remaining.

[0005] Preferably, the estimated total energy consumption is specifically: The energy consumption of the vehicle on the road section is estimated by an energy consumption calculation model based on the dynamics parameters consisting of real-time vehicle load mass, road slope and driving speed, rather than simply multiplying the driving distance and fixed energy consumption coefficient, and the total energy consumption for completing the current task is obtained by summing the energy consumptions of all road sections.

[0006] Preferably, the region urgency is specifically: The recycling service region is divided into multiple regular grids on an electronic map, and for each grid with unprocessed orders, the node urgency of the grid is calculated, the node urgency of all grids with unprocessed orders that are not covered by the candidate path is obtained, and the sum of the node urgencies of all such grids in the region is the region urgency.

[0007] Preferably, the final time cost, in particular: The candidate path is divided into several sections, for each section, the real-time average driving speed of the section and the length of the section are obtained, and the estimated driving time on the section is calculated; For each collection node that needs to stop on the path, based on the real-time inventory weight collected at the node and a preset unit weight operation efficiency, the operation time at the node is estimated; the total time is the sum of the section driving time and the node operation time; According to the set unit time cost coefficient, the total time is multiplied by the unit time cost coefficient to obtain the final time cost of executing the candidate path.

[0008] Preferably, the system state value, in particular: The linear regression fitting is performed with the regional urgency value of each network as the X-axis and the average waiting time of all unprocessed orders in the grid as the Y-axis to obtain the trend line slope; The system performs the above fitting in a fixed time window cycle to obtain a time series slope sequence; Feature extraction is performed on the time series slope sequence to obtain a feature vector containing a positive intensity index, a negative recovery index, a stability index, and a trend persistence index; The feature vector is input into a pre-set classifier to output the system state value.

[0009] Preferably, the weight of each cost, in particular: The system state identifier is quantized into a system state quantization value; The system state quantization value and the pre-set strategy bias coefficient corresponding to each cost are input into an adaptive weight calculation model; The adaptive weight calculation model is configured to: take the product of the system state quantization value and each strategy bias coefficient as the input of a natural exponential function, and normalize each exponential function output value to obtain the dynamic weight of each cost.

[0010] Preferably, the optimal collection and transportation path sequence, in particular: The weighted sum of the regional urgency cost, the energy cost, and the time cost is taken as the target comprehensive cost function, and a meta-heuristic algorithm is used to solve under the conditions of satisfying the vehicle load constraint and the task uniqueness constraint to output the optimal collection and transportation path sequence.

[0011] Preferably, the optimal collection and transportation path sequence is decomposed into navigation and operation instructions, in particular: The optimal collection and transportation path sequence is parsed into a series of continuous navigation segments and stop point operation instructions according to the order of the path passing nodes and the operation tasks of the corresponding nodes; The navigation section and the stop point operation instruction are packaged and distributed to the vehicle terminal of the corresponding recycling vehicle or the mobile application terminal of the driver through the wireless communication network.

[0012] Preferably, a renewable resource recycling method based on big data comprises the following steps: S1, through various sensing devices and data interfaces deployed in the recycling network, real-time acquisition of static basic information and vehicle operation parameters related to renewable resource recycling is performed; and standardization preprocessing of the collected raw data is performed, and finally, the multi-source data is aggregated into a structured fusion data set according to node number and time index; S2, based on the vehicle dynamics model and real-time parameters, the estimated total energy consumption is calculated, and the energy cost is calculated in combination with the energy unit price; at the same time, based on the grid area, historical data and order waiting time, the node emergency degree is calculated, and the regional urgency cost is defined accordingly; further, based on real-time traffic and node operation quantity, the total task time is estimated, and the time cost is calculated in combination with the unit time cost coefficient; finally, by analyzing the time sequence slope sequence fitted by the regional urgency and the order waiting time, a feature vector is extracted, and the system state value and specific state are output by using the pre-set classifier; S3, for receiving the system state identification, quantifying it into a specific value, and inputting it into an adaptive weight calculation model based on an exponential function and normalization processing together with a pre-set strategy bias coefficient, to dynamically generate the weight of each cost; then, an optimization function with the minimum weighted comprehensive cost as the objective is constructed, and based on the path optimization algorithm, the optimal collection and transportation path sequence of the vehicle fleet is output under the conditions of satisfying the vehicle load, task uniqueness and capacity constraints; finally, the optimal collection and transportation path sequence is decomposed into navigation and operation instructions, which are issued to the vehicle terminal through the wireless communication network.

[0013] The beneficial effects of the present application are: (1) By introducing the vehicle dynamics model for energy consumption calculation, the present application overcomes the huge error of the traditional simple mileage estimation based method, realizes the accurate prediction of fuel / electricity consumption, and provides a reliable basis for energy saving and cost reduction; at the same time, the accurate estimation of driving and operating time makes the time cost calculation more real, effectively reduces the vehicle idle and invalid driving, and greatly improves the overall operation efficiency; (2) By dynamic modeling of multiple cost factors, including energy cost, task urgency, time cost and path load distribution, and combining with normalization processing and dynamic weight driven strategy planning algorithm, the present application can generate an optimal collection and transportation path that is more consistent with real-time operation constraints. Thus, the vehicle running mileage is reduced, the attendance frequency is optimized, and the resource allocation is rationalized, thereby effectively reducing the overall operation cost and improving the resource recycling efficiency; (3) By converting service delay risk into quantifiable regional urgency cost and incorporating it into path optimization target based on dynamic weight model, the timeliness of service response and the accuracy of resource scheduling are effectively improved; through grid monitoring and time sequence trend prediction, the system can perceive the area where the service is about to be unbalanced in advance, and give priority to scheduling through cost weight in path planning, which fundamentally changes the situation of passive response to complaints, realizes proactive and forward-looking service quality guarantee, and improves customer satisfaction. BRIEF DESCRIPTION OF DRAWINGS

[0014] The application will be further described below with reference to the drawings.

[0015] Figure 1 is a principle block diagram of a renewable resource recycling system based on big data of the application.

[0016] Figure 2 is a flowchart of a renewable resource recycling method based on big data of the application. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the application.

[0018] Take a renewable resource recycling enterprise as an example. The enterprise is responsible for intelligent collection and transportation of recyclable materials in multiple communities and supermarkets in the city. The system collects multi-dimensional data including vehicle GPS track, user delivery frequency, recycling bin overflow rate, energy consumption, battery endurance, operation time efficiency, and road congestion index in real time during operation, and cooperates with the cloud big data platform through edge devices.

[0019] To solve the above technical problems, please refer to Figure 1 The application is a renewable resource recycling system based on big data, which comprises a data acquisition module, a resource recycling analysis module, and a planning and scheduling module. The data acquisition module acquires static basic information of the recycling network and vehicle operation parameters through system configuration database, Internet of Things sensing devices deployed in recycling nodes and vehicles, user terminal interface, and third-party map and energy service API. Standardized preprocessing is performed on the collected raw data, and finally multi-source data is aggregated into structured fusion data set according to node number and time index. a resource recycling analysis module, which calculates the total energy consumption based on a vehicle dynamics model and real-time parameters, and calculates the energy cost in combination with the unit price of energy; at the same time, the node urgency is calculated based on the grid area, historical data and order waiting time, and the area urgency is defined accordingly; the total task time is also estimated based on real-time traffic and node workload, and the time cost is calculated in combination with the unit time cost coefficient; finally, the time sequence slope sequence fitted by analyzing the area urgency and order waiting time is extracted to obtain the feature vector, and the system state value and specific state are output by using the pre-set classifier; a planning and scheduling module, which quantifies the system state value into a specific value based on the received system state value, and inputs the specific value and a pre-set strategy bias coefficient into an adaptive weight calculation model based on an exponential function and normalization processing to dynamically generate the weight of each cost; then, an optimization function with the minimum weighted comprehensive cost as the objective is constructed, and the optimal collection and transportation path sequence of the vehicle fleet is output by solving the optimization function based on a path optimization algorithm under the conditions of satisfying the vehicle load, task uniqueness and capacity constraints; finally, the optimal collection and transportation path sequence is decomposed into navigation and operation instructions, which are sent to the vehicle terminal through a wireless communication network; Further, in order to better introduce the technical scheme of the embodiment of the present application, based on a renewable resource recycling system based on big data, as shown in Figure 2 The embodiment of the present application provides a renewable resource recycling method based on big data, which specifically comprises the following steps: S1, real-time acquisition of static basic information and vehicle operation parameters related to renewable resource recycling through various sensing devices and data interfaces deployed in the recycling network; and standardization preprocessing of the collected raw data, and finally aggregation of the multi-source data into a structured fusion data set according to node number and time index; The S1 comprises the following steps: S11, the system obtains the static basic information of the recycling network from the pre-set database or through the management configuration interface, and the static basic information includes: the unique identifier, GPS latitude and longitude coordinate information and theoretical maximum capacity value of all resource recycling nodes (such as intelligent recycling bins, recycling stations, resource processing centers or transfer stations); the unique identifier of all resource recycling vehicles, the energy type of the vehicle, the reference rolling resistance coefficient and the air resistance coefficient of the vehicle; S12, real-time acquisition of vehicle operation parameters of the recycling network running state through Internet of Things sensing devices deployed at each recycling node, user terminal interfaces and external data interfaces; the vehicle operation parameters include: real-time vehicle load mass, rated load mass, empty vehicle weight and equipment full load state flag uploaded by the resource recycling node; Real-time location, driving speed, remaining load capacity, vehicle battery level or fuel level of the resource recycling vehicle; Weather data, traffic congestion index, order information and resource recycling area type provided by the external public interface; the resource recycling area type is a commercial area, an industrial area, a residential area, etc. S13, the acquired static basic information and vehicle operation parameters are converged and uniformly coded to form an original multi-source data set, and a preprocessing operation is performed on the original multi-source data set, including: Abnormal data detection and elimination, removing invalid, redundant and format inconsistent records; Missing value completion, missing data items are recovered by time interpolation, spatial neighbor estimation or historical mean completion method; Timestamp alignment and spatial coordinate mapping, unify the data of different sampling periods to a set time step, and perform spatial registration based on geographic coordinates; Data standardization processing, due to the huge difference in physical dimension and numerical range of different features in the multi-source data set (for example, the inventory weight is hundreds of kilograms, and the traffic index is individual digit), in order to avoid the dimension difference from being covered by the numerical magnitude in the subsequent data fusion and feature weighting process, the present application adopts the minimum-maximum normalization or Z-score standardization method to convert all features to a unified dimensionless scale; It needs to be specially pointed out that this standardization processing is intended to serve the fusion and weighting analysis of multi-source data, for the subsequent special calculation based on physical law (such as the vehicle dynamics energy consumption calculation described in S21), the system will directly call the original parameter value with physical unit; S14, feature fusion is performed on the standardized multi-source data, specifically: Data from different sources are aggregated according to node number and time index, and the historical recycling amount, equipment state and environmental features of the recycling node are weighted and integrated according to the time window to form a structured fusion data set; The fusion data set is used as input data for the subsequent recycling demand prediction model; In this embodiment, through the above steps, the system realizes efficient collection and fusion of static information and dynamic operation parameters of the recycling network, and constructs a data basis with unified format, continuous time and spatially alignable, which provides reliable data support for subsequent prediction analysis and dynamic scheduling; S2, by constructing a comprehensive analysis and prediction model, the execution cost of the recycling task in the three dimensions of energy, time and service urgency is quantitatively evaluated, and the global operation state of the system is diagnosed, which provides accurate and multi-objective quantitative input for subsequent path optimization decision; The S2 includes the following steps: S21, based on the obtained vehicle operating parameters, including the real-time load mass of the vehicle, the rated load mass and the empty mass, the real-time total mass of the vehicle is calculated; Based on the obtained real-time average driving speed of the road and the road grade, the reference rolling resistance coefficient and the air resistance coefficient of the vehicle are determined, and the reference rolling resistance coefficient and the air resistance coefficient are pre-stored in the vehicle information database; The elevation of each node of the planned path is obtained by calling the elevation service of the third-party map API, and the slope value of the road section is calculated accordingly; The above parameters are input into the vehicle dynamics energy consumption model to calculate the energy consumption of the vehicle on the road section i: Wherein, Ei is the energy consumption of the vehicle on the road section I (kJ or kWh), Di is the total length of the road section i, i=1, 2, 3…I, i is a positive integer, i represents the serial number of the road section, Fi is the rolling resistance, μ is the rolling resistance coefficient, m is the real-time total mass, g is the acceleration of gravity, θ is the slope angle, Zi is the slope resistance, Qi is the air resistance, Wherein, c is the wind resistance coefficient, A is the windward area of the vehicle, ρ is the air density, V is the speed (m / s); The above vehicle dynamics energy consumption model is based on the dynamics parameters composed of the real-time load mass of the vehicle, the road slope and the driving speed, and is estimated by the energy consumption calculation model, rather than simply multiplying the driving mileage and the fixed energy consumption coefficient; The energy consumption of all road sections is summed up to obtain the estimated total energy consumption E 总 of completing the current task; S22, the system obtains the real-time energy unit price P (such as fuel price yuan / L, or charging price yuan / kWh) of the current area from the accessed energy price service API; According to the energy type of the vehicle, the cost is calculated: If the recovered vehicle is a fuel vehicle, the system obtains the real-time fuel unit price H1 (unit: yuan / L) through the energy price interface, and calculates the final fuel energy cost based on the estimated total energy consumption E 总 According to the following formula: Wherein, L 燃油 Is the unit energy fuel equivalent of the vehicle engine (unit: kWh / L), which can be obtained from the vehicle manufacturer's calibration value or vehicle historical operation data fitting, and the final fuel energy cost calculation method is as follows: The final fuel energy cost C1 is obtained; If the recovered vehicle is a pure electric vehicle, the system obtains the real-time electricity price H2 (unit: yuan / kWh), and calculates the electricity cost according to the following formula: The final electricity energy cost C2 is obtained; If it is a hybrid vehicle, according to the preset fuel-electric drive energy distribution ratio model, the preset fuel and electric energy proportion parameters are β1 and β2, and β1+β2=1 is satisfied; the calculation method of the hybrid energy cost is: Finally, the hybrid energy cost C3 is obtained; Through the above calculation method, the energy cost evaluation result corresponding to the vehicle performing the current task can be obtained, which provides a basis for subsequent analysis; S23, dynamically assess the order backlog risk of different geographical regions, and predict the regions that need to be prioritized by analyzing the global trend data, thereby realizing the forward-looking optimization of resources; specifically including: S231, divide the recycling service area into a plurality of regular grids (such as 1kmx1km square grids) on an electronic map, and assign a unique region ID to each grid; Calculate a static basic urgency coefficient Bi for each grid, which is calculated based on: Historical average daily order density: the average number of daily orders generated in the grid in the past year; Regional type coefficient: a preset weight coefficient, for example, commercial areas and high-end residential areas are assigned higher coefficients, and industrial areas and suburban areas are assigned lower coefficients; Real-time monitoring of all generated but unprocessed recycling orders, obtaining the grid ID where the recycling order is located, and obtaining the waiting time of the order, the order waiting time being the time (unit: minutes) elapsed from the order generation to the current time, and the average waiting time being recorded as Ti; S232, for each grid i with unprocessed orders, calculate its node urgency Ui, which integrates the static basic properties and dynamic waiting conditions, and the calculation formula is as follows: Where α1 and α2 are preset normalization weight coefficients for balancing the influence of historical trends and real-time urgency; The regional urgency is defined as: in a candidate path, the sum of the node urgency of all grids with unprocessed orders that are not covered by the path, and the sum of the node urgency of all grids i in the region is the regional urgency; S233, based on the collected data, predict the total time required for the vehicle to travel along the candidate path, including road segment travel time and node operation time; specifically as follows: Road segment travel time: divide the candidate path into a plurality of road segments, for each road segment, obtain the real-time average travel speed of the road segment and the length of the road segment, and calculate the estimated travel time on the road segment; Node operation time: for each recycling node that needs to stop on the path, the operation time at the node is estimated based on the real-time inventory weight of the node and a preset unit weight operation efficiency (unit: kg / s); The total time is the sum of the road travel time and the node operation time; A unit time cost coefficient is set, which is calculated based on the historical operation data of the enterprise and is preloaded in the system database for periodic update, and the enterprise's labor cost, vehicle fixed loss and management and operation cost are analyzed; The total time obtained is multiplied by the unit time cost coefficient to obtain the final time cost of the candidate path, which is denoted as Gi; S234, taking the value of the area urgency of each network as the X axis and the average waiting time of all unprocessed orders in the grid as the Y axis, a data point is generated in a two-dimensional rectangular coordinate system; A straight line that best represents the overall trend is obtained by linear regression fitting of all grid data points using the least squares method, and its equation is Y=kX+b, where k is the slope of the trend line; The system takes a configurable fixed time window as a period to cyclically execute steps S231 to S232, thereby obtaining a slope sequence K={k1,k2,k3,...,kn} arranged in time sequence, where n is the number of historical periods counted, and the sequence constitutes a time sequence slope sequence of the system state; The setting of the collection frequency is as follows: the system automatically adjusts the collection frequency according to the variance of the historical slope sequence. When the sequence variance is greater than a preset threshold (indicating that the system is in an unstable state), the collection frequency is automatically increased, for example, shortened to every 5 minutes; when the sequence variance is less than the threshold (indicating that the system is in a stable state), the collection frequency is reduced, for example, extended to every 30 minutes; The following feature extraction and calculation are performed on the above time sequence slope sequence K: The sum of all slope values greater than 0 in the sequence is calculated and denoted as positive intensity index M1, which quantifies the cumulative intensity and total momentum of the adverse trend of waiting time extension in high urgency areas in the past period of time; The sum of all slope values less than 0 in the sequence is calculated and denoted as negative recovery index M2, which quantifies the cumulative ability and positive trend of self-repair and optimization of resource allocation of the system; The number of periods in which the slope value is equal to 0 or close to 0 in absolute value is counted and denoted as stability index M3, which reflects the persistence and frequency of the system in a stable and healthy state; The average length of the sequence in which the slope sign is continuously constant is calculated, denoted as a trend persistence index M4, for example, the average length of the paragraphs in which the consecutive positive and the consecutive negative are calculated; the value determines how much inertia a trend has once it is formed, and the stronger the persistence means that more decisive intervention is needed; S235, input the calculated feature vector M=(positive intensity index M1, negative recovery index M2, stability index M3, and trend persistence index M4) into a pre-defined fuzzy logic rule set or a trained lightweight classifier (such as decision tree, support vector machine SVM); output a clear system state value ranging from 0 to 1, and map it to the corresponding operating state according to the interval in which the value is located, as follows: the system state value is greater than 0.75, which is an imbalance aggravation alarm state; the system state value is greater than 0.5 and less than or equal to 0.75, which is a stable operation state; the system state value is greater than 0.25 and less than or equal to 0.5, which is a repair and optimization state; and the system state value is less than or equal to 0.25, which is a shock and disorder state; In this embodiment, a comprehensive prediction model integrating energy consumption quantification, regional risk assessment, and time cost analysis is constructed to provide multi-dimensional and high-precision decision input for path planning; first, the traditional simple mileage energy consumption estimation is abandoned, and a vehicle dynamics model is introduced to accurately calculate the energy cost based on real-time load, road slope, and speed; second, through a regional grid model, historical data and order waiting time are fused to dynamically evaluate the node urgency, thereby quantifying the service delay risk as a regional urgency cost; finally, the regional urgency and waiting time are analyzed in terms of time sequence slope trend, and an intelligent classifier is used to diagnose the global system state; through the above scheme, the energy cost has been transformed from rough estimation to accurate calculation, the service risk has been transformed from fuzzy judgment to forward-looking quantification, and the system has been endowed with the ability to understand the macro situation, laying a solid data foundation for subsequent adaptive optimization decision-making, and driving the entire recycling scheduling process towards intelligence and precision.

[0020] S3, according to the multi-dimensional cost and system state information, through an adaptive intelligent decision-making mechanism, a globally optimal recycling path is dynamically generated, and the planning result is converted into executable instructions to drive the vehicle in the physical world to complete the recycling task, finally realizing efficient, low-cost and rapid response resource recycling scheduling; The S3 includes the following steps: S31, according to the output system state value and the corresponding state, a state strategy mapping table is pre-set in the system to convert the state identifier into a continuous system state quantitative value; The system state quantization value is set as J, the system calls the preset mapping table to perform conversion, and an exemplary mapping relationship is as follows: when the system state is an imbalance aggravation alarm, the system state quantization value J = 2; when the system state is in repair and optimization, the system state quantization value J = 0.5; when the system state is a shock disorder, the system state quantization value J = 0; and when the system state is stable operation, the system state quantization value J = -1; Subsequently, the system calls a preset strategy bias coefficient Vn, where n represents the serial number of the cost factor, and each Vn is a strategy bias coefficient of the corresponding nth cost factor. The strategy bias coefficient is a non-negative real number and can be set by historical optimization or operation strategy, and the preferred value range is 0.5-3.0. The strategy bias coefficient reflects the importance bias of the operator to different cost items under a given system state. Based on the system state quantization value J and the strategy bias coefficient Vn, the response value Rn of the nth cost is calculated according to the following steps, where exp(·) represents the natural exponential function, i.e., the exponential function with the mathematical constant e as the base, which is used to ensure that the response has a continuous and monotonic amplification or reduction effect on the state change; the bias coefficient is included in the exponential with Vn as the coefficient, which can adjust the sensitivity of each factor to the system state quantization value J; To ensure that the obtained weight satisfies the probability distribution characteristics, the system performs normalization processing on each response value Rn, so that the sum of all weights is equal to 1. The instant weight obtained after normalization processing is used to reflect the theoretical weight distribution under the current regulation state. To avoid scheduling shock in the short-period fluctuation scenario, the system performs smoothing processing on the instant weight based on the exponential weighted moving average (EWMA) to generate a smooth output weight. This smoothing process can effectively suppress the interference of short-term abnormal data on the weight model and improve the robustness of path calculation; In the case where multiple cost factors are significantly improved at the same time, the system performs a conflict processing mechanism to prevent the weight from being concentrated on a certain type of factor and causing scheduling bias imbalance. The conflict processing includes priority correction, proportional scaling, and re-normalization steps in turn. The priority correction is used to increase the weight of the key factor according to the preset operation strategy. If the weight of a certain type of factor or factor group exceeds the set upper limit (such as 0.90) after correction, the system reduces its proportion in a proportional scaling manner, and the remaining part is redistributed to low-priority factors. Finally, the corrected weight is normalized and smoothed to ensure that the output weight is continuous, controllable, and can be directly used in the scheduling objective function. The various parameters in this step (for example, system state quantization value J, policy bias coefficient Vn) can be obtained through historical data back fitting, expert experience setting or online learning mechanism, and support adjustment by the operator through the management interface during system operation; This mechanism ensures the reproducibility, configurability and stability of weight calculation, and thus makes the path planning adaptive and implementable in different system states; The system state quantization value and the system preset policy bias coefficient Vi are input into the adaptive weight calculation model to dynamically generate the weight of each cost, wherein the weight calculation model is: Where Wn is the dynamic weight of the nth cost, for example, n=1 corresponds to the weight of the regional urgency cost, n=2 corresponds to the weight of the energy cost, and n=3 corresponds to the weight of the time cost; Vi is the policy bias coefficient of the ith cost, which is a fixed preset value, and represents the core scheduling strategy; Where m in Vm is a loop variable, which represents that all weight categories need to be calculated and summed when calculating the denominator of the formula, and Vi and Vm are the same set of policy bias coefficients; The policy bias coefficient Vi is a set of preset fixed parameters, and its value is determined according to the strategy target of the renewable resource recycling method under different system states; S32, a target function for path optimization, that is, a comprehensive cost function, is constructed, specifically: Where C1 is the calculated fuel energy cost, or the power energy cost C2 or the mixed energy cost C3, and different energy costs are selected according to different vehicle types; S33, the constructed comprehensive cost function is input into the path optimization algorithm (such as genetic algorithm, tabu search algorithm, etc.) together with the vehicle position, load, recycling point inventory and all other constraint conditions for calculation; The optimization process needs to meet the following core constraints: Vehicle load constraint: the sum of recycling amounts of all nodes on the path cannot exceed the rated load of the vehicle; Task uniqueness constraint: each task is completed at most once by one vehicle; Vehicle capacity constraint: the actual load of the vehicle at any point on the path cannot exceed its capacity; Time window constraint (if applicable): the task needs to be completed within the specified time range; Output: After the algorithm is calculated, a set of optimal collection and transportation path sequences for the entire vehicle fleet is output, which ensures that the global comprehensive cost is the lowest under the current system state; The termination condition is that the number of iterations reaches 200 times; S34: task dispatching and instruction issuing; The system decomposes the optimal path sequence generated in S33 into specific executable navigation instructions and operation tasks, and issues these instructions and tasks to the vehicle terminal of the corresponding recycling vehicle or the mobile application APP of the driver through the wireless communication network, to guide it to complete the recycling operation. In the embodiment, the system state value is quantified as a specific numerical value, which is input into a calculation model based on a natural exponential function and normalization processing together with a preset strategy bias coefficient, to dynamically generate real-time weights of regional urgency, energy and time cost; subsequently, the system constructs a comprehensive cost function based on the dynamic weights, and uses a path optimization algorithm to solve a globally optimal collection and transportation path sequence under the constraints of load, task uniqueness and other realistic constraints; finally, the system automatically decomposes the optimal path into specific navigation and operation instructions, and issues them to the vehicle terminal for execution; through the scheme of this step, the scheduling strategy is upgraded from static and fixed to dynamic and adaptive, which can intelligently adjust and optimize the target according to the real-time state of the system, ensures that the path planning result has global optimality and practical feasibility, completes the precise closed-loop control from digital decision to physical execution, and finally significantly improves the overall efficiency, economy and response speed of the recycling operation.

[0021] The present application constructs an integrated intelligent solution from perception to decision to execution, aiming at the core pain points of inaccurate cost estimation, rigid strategy and decision-making divorced from reality in traditional recycling scheduling; the core idea is that: first, the three costs of energy, time and service urgency are accurately quantified through vehicle dynamics model and regional grid analysis; second, the adaptive weight decision mechanism can dynamically adjust the cost weight according to the real-time state of the system (such as service imbalance alarm), to realize intelligent switching of scheduling strategy; finally, the globally optimal path is generated and automatically issued for execution under the strict satisfaction of actual operation constraints; the present application realizes the leap from static mechanical planning to dynamic intelligent control, and significantly improves the economy, efficiency and service quality of recycling operation.

[0022] The above content is only an example and description of the structure of the present application, and those skilled in the art can make various modifications or supplements or use similar ways to replace, as long as the modifications or supplements do not deviate from the structure of the present application or exceed the scope defined by the present claims, and should belong to the protection scope of the present application.

Claims

1. A big data based renewable resource recycling system, characterized in that, Comprise; The data acquisition module obtains the static basic information of the recycling network and the vehicle operation parameters through the system configuration database, the Internet of Things sensing device deployed in the recycling node and the vehicle, the user terminal interface, and the third-party map and energy service API, and performs standardization preprocessing on the collected raw data, and finally aggregates the multi-source data into structured fusion data set according to the node number and time index; The resource recycling analysis module estimates the total energy consumption based on the vehicle dynamics model and real-time parameters, calculates the energy cost in combination with the unit price of energy, calculates the node urgency based on the grid area, historical data and order waiting time, defines the area urgency accordingly, estimates the total task time based on real-time traffic and node workload, calculates the final time cost in combination with the unit time cost coefficient, and finally extracts the feature vector by analyzing the time sequence slope sequence fitted by the area urgency and order waiting time, and outputs the system state value and specific state using the pre-set classifier; The planning and scheduling module quantifies the system state value into a specific value based on the received system state value, inputs it into the adaptive weight calculation model based on the exponential function and normalization processing together with the pre-set strategy bias coefficient, and dynamically generates the weight of each cost; Subsequently, an optimization function with the minimum weighted comprehensive cost as the target is constructed, and based on the path optimization algorithm, the optimal collection and transportation path sequence of the vehicle fleet is output under the conditions of satisfying the vehicle load, task uniqueness and capacity constraints; finally, the path sequence is decomposed into navigation and operation instructions and issued to the vehicle terminal through the wireless communication network.

2. The renewable resource recycling system based on big data according to claim 1, characterized in that, The static basic information and vehicle operation parameters specifically include: The static basic information includes the unique identifier of all resource recycling nodes, GPS latitude and longitude coordinate information, and the theoretical maximum capacity value of the recycling node, the unique identifier of all resource recycling vehicles, the energy type of the vehicle, the baseline rolling resistance coefficient and air resistance coefficient of the vehicle; The vehicle operation parameters include the real-time vehicle load mass, rated load mass and empty vehicle weight uploaded by the resource recycling node, equipment full load state flag, real-time position, driving speed, remaining load capacity, vehicle battery power or fuel remaining.

3. The renewable resource recycling system based on big data according to claim 1, characterized in that, The estimated total energy consumption specifically is: Based on the dynamics parameters composed of real-time vehicle load mass, road slope and driving speed, the energy consumption is estimated by the energy consumption calculation model, the energy consumption of the vehicle on the road section is obtained, and the sum of the energy consumption of all road sections is obtained to obtain the estimated total energy consumption for completing the current task.

4. The renewable resource recycling system based on big data according to claim 1, characterized in that, The area urgency specifically is: The recycling service area is divided into multiple regular grids on the electronic map, for each grid with unprocessed orders, the node urgency of the grid is calculated, the node urgency of all grids with unprocessed orders that are not covered by the candidate path is obtained, and the sum of the node urgencies of all such grids in the area is the area urgency.

5. The big data based renewable resource recycling system of claim 1, wherein, The final time cost specifically is: The candidate path is divided into several road sections, for each road section, the real-time average driving speed of the road section and the length of the road section are obtained, and the estimated driving time on the road section is calculated; For each recycling node that needs to stop on the path, the working time at the node is estimated based on the real-time inventory weight of the node and a preset unit weight working efficiency; The total time is the sum of the road travel time and the node working time; According to the set unit time cost coefficient, the total time is multiplied by the unit time cost coefficient to obtain the final time cost of the execution candidate path.

6. The big data based renewable resource recycling system of claim 1, wherein, The system state value is specifically: A trend line slope is obtained by linear regression fitting with the regional urgency value of each network as the X-axis and the average order waiting time of all unprocessed orders in the grid as the Y-axis; The system performs the above fitting in a fixed time window cycle to obtain a time series slope sequence; Feature extraction is performed on the time series slope sequence to obtain a feature vector containing a positive intensity index, a negative recovery index, a stability index, and a trend persistence index; The feature vector is input into a preset classifier to output the system state value.

7. The system of claim 1, wherein, The weight of each cost is specifically: The system state identifier is quantized into a system state quantization value; The system state quantization value and the preset strategy bias coefficient corresponding to each cost are input into an adaptive weight calculation model; The adaptive weight calculation model is configured to take the product of the system state quantization value and the strategy bias coefficient as the input of a natural exponential function, and normalize the output values of each exponential function to obtain the dynamic weight of each cost.

8. The big data based renewable resource recycling system of claim 1, wherein, The optimal collection and transportation path sequence is specifically: A comprehensive cost function with the weighted sum of the regional urgency cost, the energy cost, and the time cost as the target is used to solve the optimal collection and transportation path sequence under the conditions of satisfying the vehicle load constraint and the task uniqueness constraint.

9. The big data based renewable resource recycling system of claim 8, wherein, The optimal collection and transportation path sequence is decomposed into navigation and operation instructions, specifically: The optimal collection and transportation path sequence is parsed into a series of continuous navigation segments and stop point operation instructions according to the order of the nodes on the path and the operation tasks of the corresponding nodes; The navigation segments and stop point operation instructions are packaged and distributed to the vehicle-mounted terminals of the corresponding recycling vehicles or the mobile application terminals of the drivers through a wireless communication network. 10.A method for recycling renewable resources based on big data, characterized in that, The system of any one of the above claims 1-9 comprises the following steps: S1, real-time acquisition of static basic information and vehicle operation parameters related to renewable resource recycling through various sensing devices and data interfaces deployed in the recycling network; and standardization preprocessing of the collected raw data, finally aggregating the multi-source data into a structured fusion data set according to node number and time index; S2, based on the vehicle dynamics model and real-time parameters, the total energy consumption is calculated and estimated, and the energy cost is calculated combined with the energy unit price; at the same time, based on the grid area, historical data and order waiting time, the node urgency is calculated, and the regional urgency cost is defined accordingly; the total task time is also estimated based on real-time traffic and node operation amount, and the time cost is calculated combined with the unit time cost coefficient; finally, the time series slope sequence is fitted by analyzing the regional urgency and order waiting time, the feature vector is extracted, and the system state value and specific state are output by using the preset classifier. S3, for receiving the system state identification, quantifying it into a specific value, and inputting it into an adaptive weight calculation model based on an exponential function and normalization processing together with a preset strategy bias coefficient to dynamically generate the weight of each cost; then, an optimization function with the goal of minimizing the weighted comprehensive cost is constructed, and based on a path optimization algorithm, the solution is performed under the condition of satisfying the vehicle load, task uniqueness and capacity constraints, and the optimal collection and transportation path sequence of the vehicle fleet is output; finally, the optimal collection and transportation path sequence is decomposed into navigation and operation instructions and issued to the vehicle terminal through a wireless communication network.

Images