Smart park collaborative service system based on cloud computing and big data

Abstract

The invention discloses a smart park collaborative service system based on cloud computing and big data, and relates to the technical field of smart parks. Which comprises a data security and privacy protection module, and is characterized in that a data acquisition module is used for acquiring multi-source data of the smart park in real time; the data processing module is used for constructing a data fusion and collaboration model, performing feature extraction and fusion on the multi-source data of the smart park, performing collaborative analysis based on big data, generating collaborative decision suggestions, and realizing data intercommunication and sharing; and the intelligent service and optimization module is used for constructing an intelligent service model, automatically generating services based on service requests, tracking and feeding back the service progress in real time, optimizing and updating the model and improving the service response efficiency. According to the invention, by constructing the intelligent service model, analyzing the multi-source service request in real time, automatically generating the service and optimizing and updating the model, the service response speed and the user service experience are continuously improved, and the user loss risk is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of smart park, in particular to a smart park collaborative service system based on cloud computing and big data. BACKGROUND

[0002] The smart park service system is a key carrier for the transformation and upgrading of traditional parks in the digital economy era. It is born out of the need to solve the real pain points of park management efficiency, service experience, cost control, and security protection, relying on information technologies such as the Internet of Things, cloud computing, big data, and artificial intelligence, aiming to create a modern park ecosystem that is efficient in management, convenient in service, pleasant in environment, active in industry, and green and sustainable.

[0003] For example, patent publication No. CN115423664A, entitled "Smart Park Internet of Things Comprehensive Service System", includes an energy management system, an access control system, a video surveillance system, a fire alarm system, an image processing system, an intelligent cleaning system, a processor, and a display system. Each system is connected to the processor. The invention makes the park highly intelligent and facilitates centralized management.

[0004] However, the above and similar technical solutions lack deep analysis of data, which may result in incomplete decision-making suggestions, resource allocation lag, slow response to user service requests, poor user experience, reduced user satisfaction, and increased risk of user loss. SUMMARY

[0005] The present application aims to provide a smart park collaborative service system based on cloud computing and big data to solve the problems raised in the background.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical solution: a smart park collaborative service system based on cloud computing and big data, comprising a data security and privacy protection module, characterized in that:

[0007] Data acquisition module: for real-time acquisition of multi-source data of smart park;

[0008] Data processing module: for building data fusion and collaboration model, extracting and fusing features of multi-source data of smart park, and based on big data, for collaborative analysis, generating collaborative decision-making suggestions, realizing data intercommunication and sharing;

[0009] Smart service and optimization module: for building a smart service model, automatically generating services based on service requests, and tracking and feeding back service progress in real time, optimizing and updating the model, and improving service response efficiency.

[0010] Further, the method for creating the data fusion and collaboration model comprises:

[0011] Data processing: standardize and clean the collected multi-source data of the smart park, and perform coordinate conversion and timestamp synchronization according to the unified space-time reference to obtain the processed multi-source data of the smart park;

[0012] Feature fusion: based on the space-time-semantic association knowledge network, the processed multi-source data of the smart park is correlated across modalities to establish a fusion database;

[0013] Decision coordination: based on big data, the data in the fusion database are analyzed to mine the association relationships between the data, and collaborative decision suggestions are generated.

[0014] Further, the method for constructing the space-time-semantic association knowledge network comprises:

[0015] Data annotation: giving the processed multi-source data of the smart park preliminary semantic labels;

[0016] Geospatial association construction: constructing a spatial geographic model of the smart park according to the processed multi-source data of the smart park; calculating the spatial relationships between different spatial elements, including distance, adjacency and inclusion;

[0017] Time element extraction: analyzing the processed multi-source data of the smart park to extract time rules and mine time patterns to obtain space-time elements;

[0018] Semantic association knowledge extraction and fusion: extracting the semantic knowledge of the smart park from the processed multi-source data of the smart park and related literature, and constructing a semantic knowledge base; based on the semantic knowledge base, correlating the processed multi-source data of the smart park with the semantic knowledge;

[0019] Network construction: taking the processed multi-source data of the smart park, space-time elements, semantic knowledge and spatial geographic model and their spatial relationships as nodes, and the association relationships between the nodes and the spatial relationship weights as edges, to construct a space-time-semantic association knowledge network

[0020] Further, the step of generating collaborative decision suggestions comprises:

[0021] Data analysis: performing multi-dimensional analysis on the data in the fusion database, including time dimension, space dimension and modal dimension, to obtain data analysis results;

[0022] Association rule mining: according to the data analysis results, mining the association relationships between the data in the fusion database; and according to the association relationships, predicting future data trends and event occurrences to obtain prediction results;

[0023] Generating decision suggestions: constructing decision scenarios, determining decision indicators and influencing factors; according to the association relationships and prediction results, generating collaborative decision suggestions for the corresponding scenarios.

[0024] Further, the feature fusion method comprises:

[0025] Cross-modal feature association: splicing the processed smart park data to generate a basic feature vector; weighting and aggregating multi-modal features to obtain a multi-modal feature vector;

[0026] Dimensionality reduction and optimization: dimensionality reduction processing of the spatio-temporal-semantic association knowledge network and the multi-modal feature vector to obtain a reduced feature vector, and adjusting the feature weight to establish a fusion database.

[0027] Further, the cross-modal feature association method comprises:

[0028] Splicing: concatenating the feature vectors of different modalities aligned within the same spatio-temporal unit to obtain a basic feature vector;

[0029] Computing feature weights: automatically learning the specific content of the basic feature vector to obtain dynamic weights for each feature dimension, highlighting key information;

[0030] Weighted aggregation: weighting and summing the computed feature weights and the basic feature vector to obtain a multi-modal feature vector.

[0031] Further, the method of adjusting feature weights comprises:

[0032] Normalized reduced dimension feature: normalizing the reduced feature vector and reallocating the initial weight according to the feature variance contribution;

[0033] Computing dynamic weights: analyzing the semantic association between multi-modal features to compute dynamic weight coefficients;

[0034] Similarity-driven adjustment: computing the similarity matrix between feature vectors and adjusting the feature weights according to the association strength.

[0035] Further, the method of creating a smart service model comprises:

[0036] Service function integration: integrating various service function modules;

[0037] Process definition: defining the processing flow of various services, including task allocation, processing time limit, and flow path;

[0038] Information input: user inputs service request;

[0039] Service execution: based on the processing flow, according to the service request, judging the service type, automatically generating the service, and transferring to the related service;

[0040] Service feedback and optimization: real-time tracking of service request processing progress and feedback to users; service ends, push service evaluation survey to users, analyze collected service evaluation data, and optimize processing flow according to analysis results.

[0041] Further, the service execution step comprises:

[0042] Determine service type: analyze user input service request, identify service type, and match predefined processing flow;

[0043] Service generation: according to the processing flow, the service task is assigned to the corresponding processing node, and the processing progress data is synchronized to the fusion database in real time.

[0044] Further, the service feedback and optimization includes the processing of service timeout and rule conflict, and the processing method includes: automatically triggering the early warning mechanism and switching to the artificial review node, informing the user of the service timeout reason and solution, and updating the fusion database.

[0045] Compared with the prior art, the beneficial effects of the present application are:

[0046] A smart park collaborative service system based on cloud computing and big data, by building a data fusion and collaboration model, fusing multi-source data, and based on big data, deeply analyzing and predicting the data, mastering the real-time state and use rules of various resources in the park, so that the park resources can be reasonably allocated, realizing the leap from data interconnection to intelligent decision-making, and improving the accuracy of park management.

[0047] At the same time, by building a smart service model, real-time analysis of multi-source service requests, automatic generation of services, and optimization and update of the model, continuously improving service response speed and user service experience, and reducing user churn risk. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 The system module diagram of the present application;

[0049] Figure 2 The data fusion and collaboration model of the present application;

[0050] Figure 3 The spatio-temporal-semantic correlation knowledge network construction method of the present application;

[0051] Figure 4 The smart service model of the present application. DETAILED DESCRIPTION

[0052] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work are within the protection scope of the present application.

[0053] As shown in Figure 1 The present application provides a technical solution: a smart park collaborative service system based on cloud computing and big data, comprising:

[0054] A data acquisition module is configured to acquire real-time smart park multi-source data.

[0055] It should be noted that the acquired smart park multi-source data includes user basic data, facility equipment data, service and feedback data, vehicle data, enterprise data, and is transmitted to a cloud data center in real time. The user basic data is obtained through a face / fingerprint recognition machine, a mobile terminal APP / applet, and a WiFi probe. The facility equipment data is obtained through an IOT sensor, a PLC system, a device manufacturer API, and a smart electricity / water meter. The service and feedback data is obtained through a mobile terminal, a smart terminal, and an API gateway log. The vehicle data is obtained through an entrance camera, a geomagnetic sensor, GPS positioning, and ETC. The enterprise data is obtained through an API connected to an enterprise OA system.

[0056] A data processing module is configured to construct a data fusion and collaboration model, extract and fuse features of the smart park multi-source data, and perform collaborative analysis based on big data to generate collaborative decision suggestions and realize data intercommunication and sharing.

[0057] As shown in Figure 2 The present application provides a data fusion and collaboration model.

[0058] Specifically,

[0059] H1: Data processing: standardizing and cleaning the acquired smart park multi-source data, and performing coordinate conversion and time stamp synchronization according to a unified space-time reference to obtain processed smart park multi-source data.

[0060] H2: Feature fusion: based on a space-time-semantic association knowledge network, correlating the processed smart park multi-source data across modalities to establish a fusion database.

[0061] H3: Decision collaboration: based on big data, analyzing the data in the fusion database, mining the correlation between the data, and generating collaborative decision suggestions.

[0062] The feature fusion method comprises: cross-modal feature association: splicing the processed smart park data to generate a basic feature vector; weighted aggregation of multi-modal features to obtain a multi-modal feature vector. Dimension reduction and optimization: dimension reduction processing of the spatio-temporal-semantic association knowledge network and the multi-modal feature vector to obtain a reduced feature vector, and adjusting the feature weight to establish a fusion database.

[0063] The specific method of cross-modal feature association comprises: splicing, connecting the head and tail of the feature vectors of different modalities aligned in the same space-time unit to obtain a basic feature vector; calculating the feature weight, automatically learning the specific content of the basic feature vector to obtain the dynamic weight of each feature dimension, and highlighting the key information; weighted aggregation, weighted sum of the calculated feature weight and the basic feature vector to obtain a multi-modal feature vector.

[0064] The method of adjusting the feature weight comprises: normalizing the reduced dimension feature, standardizing the reduced feature vector, and reallocating the initial weight according to the feature variance contribution; calculating the dynamic weight, analyzing the semantic association between multi-modal features, and calculating the dynamic weight coefficient; similarity driven adjustment, calculating the similarity matrix between feature vectors, and adjusting the feature weight according to the correlation strength.

[0065] The step of generating collaborative decision suggestions comprises: data analysis: multi-dimensional analysis of data in the fusion database, including time dimension, space dimension, and modal dimension, to obtain data analysis results; association rule mining: mining the association relationship between data in the fusion database according to the data analysis results; predicting future data trends and events according to the association relationship to obtain prediction results; generating decision suggestions: constructing a decision scenario, determining decision indicators and influencing factors; generating collaborative decision suggestions for corresponding scenarios according to the association relationship and prediction results.

[0066] It should be noted that data processing: using ETL tools to standardize and clean the collected multi-source data of the smart park; for abnormal values, setting threshold values based on park business rules for automatic correction; for missing values, using KNN algorithm for multivariate interpolation; then establishing a master data management module to unify field definitions, such as standardizing customer names to Unicode encoding. Through Gauss-Krueger projection, local coordinate system is converted into WGS84 standard, with an error of within ±0.5 meters; then deploy NTP server cluster to ensure that the timestamp deviation is less than 50ms.

[0067] Feature fusion: First, cross-modal feature correlation is performed: feature splicing, different modal feature vectors aligned within the same space-time unit, such as video extracted CNN features, sensor time series data, and hand-tail connection to form a basic feature vector; use the torch.cat function of PyTorch to realize vector merging. Note that before splicing, make sure that the different modal feature vectors are spatio-temporally aligned and the dimensions are unified (all feature vectors are standardized to the same length). Calculate feature weights, use cross-modal attention mechanism, automatically assign weights through learnable parameter matrix, such as using multi-head self-attention layer to calculate dynamic weight coefficients for each feature dimension, highlighting key information such as energy peak related features. Weighted aggregation, use PyTorch's broadcast mechanism to perform weighted summation of dynamic weights and basic feature vectors to generate multi-modal feature vectors. Note that the weights need to be normalized to avoid bias caused by feature scale differences.

[0068] Then, dimension reduction and optimization: dimension reduction processing is performed on the spatio-temporal-semantic correlation knowledge network stored in the Neo4j graph database and the multi-modal feature vector, and the principal component analysis is used to decompose the feature value to realize dimension reduction, and the principal components with 95% variance contribution are retained. Note that the feature dimension after dimension reduction needs to be lower than 30% of the original data to ensure that the information loss rate is ≤5%. Normalize the dimension reduction features, apply Z-score standardization to process the dimension reduction features, and redistribute the initial weights according to the feature variance contribution, with higher variance resulting in higher feature weights. Calculate dynamic weights, analyze the semantic correlation between multi-modal feature vectors, such as device-energy relationship, and output dynamic weight coefficients; then calculate node correlation through GAT. Similarity-driven adjustment, calculate the cosine similarity matrix between feature vectors, and based on the correlation strength, adjust the weights. Weight adjustment methods include: when the similarity is greater than 0.7, the feature weights are superimposed; when the similarity is between 0.3 and 0.7, the weights are conditionally weighted; when the similarity is less than 0.3, the weights are filtered.

[0069] Finally, the establishment of the fusion database: database type, heterogeneous database based on feature layer fusion, supporting hierarchical storage: hot data is stored in Redis; warm data is stored in InfluxDB time series database; cold data is archived to AWS S3. The hot data is real-time features, the warm data is dimension reduction features, and the cold data is original smart park multi-source data. Data structure, use graph structure to store spatio-temporal-semantic correlation, such as entity relationship "camera-human flow density-BIM area", feature vector is stored in Float32, supporting efficient similarity query. Function implementation, provide unified RESTful API interface, support real-time access by multiple business systems; have dynamic weight indexing mechanism, accelerate key feature retrieval. The fusion database solves the problem of multi-source data island in smart park through cross-modal correlation and optimized dimension reduction, providing high-density feature support for decision-making collaboration.

[0070] Decision synergy: data analysis, including: time dimension analysis, using STL decomposition and Prophe model to identify the periodicity and trend characteristics of the data, predicting the future 72-hour data trend through the LSTM network, the time alignment accuracy needs to reach millisecond level; the abnormal value is repaired by cubic spline interpolation, and the sliding window size is 1 hour; spatial dimension analysis, based on the construction of Geohash coded spatial index, combined with DBSCAN algorithm, identify device-intensive area and personnel activity hotspot; GIS data projection is unified in CGCS2000 coordinate system; modal dimension analysis, using multi-modal Transformer model for feature alignment, calculating feature similarity through cross-attention mechanism, threshold ≥0.7. Association rule mining, using FP-Growth algorithm to mine frequent item sets, such as "peak period → elevator congestion + energy consumption surge". Through PC algorithm to distinguish correlation and causality, exclude pseudo-association; then through ST-GNN model to fuse graph convolution and LSTM, predict the probability of event occurrence; the prediction result needs to be verified by Bootstrap sampling, confidence ≥90%; based on Apriori algorithm to generate association rules, real-time update to Neo4j graph database. Decision suggestion generation, using digital twin engine to load BIM and IOT data; determine the index weight through AHP hierarchical analysis method; based on PPO algorithm to simulate different strategy effects, output Pareto optimal scheme; emergency type decision response time less than 3 seconds, resource allocation type decision needs to include three alternative schemes. This method realizes the intelligent transformation from data to action through the closed-loop process of multi-dimensional analysis-association mining-dynamic decision.

[0071] As shown in Figure 3 The application provides a construction method of a space-time-semantic association knowledge network.

[0072] Specifically:

[0073] Data annotation: giving preliminary semantic labels to the processed smart park multi-source data;

[0074] Geospatial association construction: constructing a smart park spatial geographic model according to the processed smart park multi-source data; calculating the spatial relationship between different spatial elements, including distance, adjacency and inclusion;

[0075] Space-time element extraction: analyzing the processed smart park multi-source data, extracting time rules and mining time patterns to obtain space-time elements;

[0076] Semantic association knowledge extraction and fusion: extracting smart park semantic knowledge from the processed smart park multi-source data and related literature materials, and constructing a semantic knowledge base; based on the semantic knowledge base, associating the processed smart park multi-source data with semantic knowledge;

[0077] Network construction: Based on the processed multi-source data of smart park, spatio-temporal elements, semantic knowledge and spatial geographic model and their spatial relationships as nodes, the association between nodes and spatial relationship weights as edges, a spatio-temporal-semantic association knowledge network is constructed.

[0078] It should be noted that data annotation is for text data such as park announcements and equipment maintenance records. The sentence is split using a word segmentation tool (such as Jieba), and key entities are identified through part-of-speech tagging (such as NLTK's POS Tagger), such as Building 3 Elevator and Fire Hydrant A. Based on the BERT model for semantic understanding, the data is annotated to the category, such as equipment failure and personnel access. For video / image data, use the YOLO target detection algorithm to identify entities such as pedestrians, vehicles, and fire access, and annotate their attributes such as pedestrian-employee and vehicle-foreign vehicle. Based on pre-set business rules such as park management specifications, automatically annotate the semantics of structured data such as "card swiping time + personnel ID" in access control records, and automatically annotate "employee Zhang San + 8:40 access to the park". Annotation rules: Follow the entity-attribute-value model, such as device-status-normal and area-function-office area; unify the semantic tag dictionary to ensure consistency of the same class of entities / relationships, such as adjacent areas not mixed with adjacent areas.

[0079] Geospatial association construction is based on the processed multi-source data of smart park to construct a three-dimensional spatial model of the park. Digital modeling of spatial elements is achieved through ArcGIS tools, such as building coordinates, road alignment, and equipment installation location. The straight-line / actual path distance between two spatial elements is calculated based on the Euclidean distance formula. Through topological analysis algorithms, it is determined whether the elements share boundaries, such as Building 2 and Building 3 being adjacent, and the fire access being adjacent to Building 2 stairs. Based on the spatial inclusion algorithm, the element affiliation is determined, such as Conference Room A being included in Building 3 and the parking lot containing 120 parking spaces. Spatial relationship rules: Follow the topological relationship theory, i.e. the topological rules of points, lines, and surfaces, such as surface-surface elements cannot overlap (except for special areas) and passages must be adjacent to at least two areas. Combined with the actual layout rules of the park, such as equipment being located within the area it belongs to.

[0080] Temporal-spatial element extraction, temporal features such as time points, time periods, and periodic regularities are extracted from the processed smart park multi-source data and are associated with spatial entities to form temporal-spatial elements. Specifically, for time series data such as sensors and access control, sliding window algorithm is used to extract time points and time periods; through LSTM time series model, regularities such as device A needing maintenance every 30 days are mined; for unstructured data such as device maintenance record texts, spaCy's time parser is used to extract timestamps and associate them with corresponding spatial entities. Time rules: time granularity is standardized, and "year-month-day-hour-minute-second" format is used, and absolute time and relative time are distinguished; according to periodicity, it is divided into "one-time event", "periodic event" and "trend event". The absolute time refers to the fixed time coordinate point based on the Gregorian calendar, which represents the physical time point when the event actually occurs; the relative time refers to the time offset based on the reference event, which represents the time sequence relationship between events rather than the specific time. By strictly distinguishing absolute time and relative time, the system can support accurate time point positioning (device failure instant) and complex time sequence relationship reasoning (such as "30 minutes after power failure, standby generator automatically starts").

[0081] Semantic association knowledge extraction and fusion, domain knowledge such as device function, management rule, and event causal relationship is extracted from the processed smart park multi-source data and related literature, and is associated with temporal-spatial elements to form a structured semantic knowledge base. Specifically, based on NER model, domain entities such as smart meter, security system, and emergency plan are extracted from related literature; through remote supervision algorithm, semantic relationships between entities such as "smart meter-monitors-electricity consumption" and "security system-associates-camera" are extracted; then event extraction is performed to identify key events and elements such as "device failure event, subject: elevator A, time: 2025-07-24-15-30-20, reason: motor damage"; entity ambiguity is solved by calculating the cosine similarity of word vectors, such as "Building No. 3" and "Building No. 3" are the same entity; through Neo4j, multi-source semantic knowledge is fused into a unified knowledge base to avoid duplication or conflict. Semantic association rules: based on domain ontology, semantic relationships such as "device-belongs-to-area" and "event-triggers-corresponding measures" are defined; logical reasoning rules are followed, such as "if device A belongs to area B, and area B is contained in park C, then device A belongs to park C".

[0082] Network construction, with the processed smart park multi-source data, space-time elements, semantic knowledge and space geographic model and their spatial relationships as nodes, the association between nodes and the weight of spatial relationships as edges, a space-time-semantic association knowledge network is constructed. Specifically, the nodes and edges are stored as a graph structure with Neo4j graph database as the carrier; the edges are assigned weights to quantify the association strength, such as the weight of "device A and device B belonging to the same maintenance group" is higher than that of "device A and device C being far away". Network construction rules: node uniqueness, only one node is retained for the same entity / element in the network, which is ensured by ID mapping; relationship type needs to be standardized for subsequent query and reasoning, such as "spatial-distance", "time-chronologically", "semantic-causal". The weight of the edge is directly used as the weight value of the edge by using Euclidean distance to measure the similarity of node attributes. The constructed space-time-semantic association knowledge network is stored in a graph database, such as Neo4j.

[0083] Intelligent service and optimization module: used for constructing an intelligent service model, automatically generating services based on service requests, tracking and feeding back service progress in real time, optimizing and updating the model, and improving service response efficiency.

[0084] As shown in Figure 4 The present application provides an intelligent service model;

[0085] Specifically:

[0086] S1: service function integration: integrating various service function modules;

[0087] S2: process definition: defining the processing flow of various services, including task allocation, processing time limit and flow path;

[0088] S3: information input: user inputs service request;

[0089] S4: service execution: based on the processing flow, according to the service request, judging the service type, automatically generating the service, and transferring to the related service;

[0090] S5: service feedback and optimization: tracking the service request processing progress in real time, and feeding back to the user; after the service is completed, service evaluation survey is pushed to the user, and the collected service evaluation data is analyzed, and the processing flow is optimized according to the analysis result.

[0091] The service execution step includes: judging the service type: analyzing the user input service request, identifying the service type, and matching the predefined processing flow; service generation: according to the processing flow, the service task is assigned to the corresponding processing node, and the processing progress data is synchronized to the fusion database in real time.

[0092] The service feedback and optimization includes the processing of service timeout and rule conflict, and the processing method includes automatically triggering an early warning mechanism and switching to a manual review node, notifying the user of the service timeout reason and solution, and updating the fusion database.

[0093] It should be noted that the service function integration: integrated APP / miniprogram / smart terminal, using micro-frontend technology to realize modular function loading, users access all services through single sign-on. The reception and classification of service requests, service requests can enter the intelligent service system through various channels, such as APP, miniprogram, self-service terminal, etc. The system automatically records the basic information of the service request, including request time, request source, requester identity information, contact information, etc.; based on the content and nature of the service request, use NLP technology to automatically classify the service request, such as security service class, including monitoring video viewing application, access control fault repair, security time report; energy service class, including power failure repair, water anomaly report, energy use consultation; property service class: including housing repair application, public area cleaning complaint, greening maintenance suggestion; enterprise service class: enterprise entry related procedure consultation, service demand change application, enterprise activity organization assistance. For service requests that cannot be accurately classified automatically, mark them as "to be manually classified" and assign them to customer service personnel for manual judgment and classification.

[0094] Process definition, task allocation rules: based on service type allocation: security service class, assigned to security management department or related professional and technical personnel, such as monitoring video viewing application assigned to monitoring center staff, access control fault repair assigned to access control system maintenance personnel; energy service class, power failure repair assigned to power maintenance team, water anomaly report assigned to water management department, energy use consultation assigned to energy service officer; property service class, housing repair application assigned to repair engineering team, public area cleaning complaint assigned to cleaning team leader, greening maintenance suggestion assigned to greening maintenance department; enterprise service class, enterprise entry related procedure consultation assigned to enterprise service officer, service demand change application assigned to corresponding service manager, enterprise activity organization assistance assigned to activity planning team. Based on geographical location allocation: for service requests involving specific areas in the park, such as air conditioning fault repair of a certain office building, according to the location information provided in the service request, the task is allocated to the relevant service team and staff responsible for the area; the system can combine the map information of the park and the jurisdiction area range of each service team to automatically determine the allocation object of the task. Based on workload allocation: the system monitors the task load of each service team and staff in real time, including the number of current unfinished tasks, the emergency degree of task urgency and the estimated completion time; for new service requests, under the premise of meeting the above service type and geographical location, preferentially allocate to the service team or staff with relatively light workload, to realize the balanced allocation of tasks and improve the overall service efficiency.

[0095] Special rule allocation: for high-priority service requests, such as those involving security risks, major facility failures, or important customer service needs, the system uses priority allocation rules to immediately assign tasks to senior technicians or specialized emergency teams in the relevant service team, and marks them as urgent tasks requiring priority handling; for users with specific service requirements or historical service records, such as long-term corporate clients or park users with special needs, the system will allocate tasks to dedicated staff familiar with the client based on pre-set customer preferences and service agreements to improve personalized and high-quality service.

[0096] Processing time limit setting: according to different service types, corresponding basic processing time limits are formulated, such as for security services, monitoring video viewing applications should be responded to within 15 minutes, providing video viewing services; access control failure repair should arrive on site for repair within 30 minutes; security event reports should initiate emergency response procedures within 10 minutes. For energy services, power failure repair should restore power within 3 hours, with specific time limits set according to fault severity and repair difficulty levels; water anomaly reporting should identify abnormal causes and take preliminary treatment measures within 2 hours; energy use consultation should be responded to within 20 minutes. For property services, housing repair applications set different processing time limits according to the urgency and complexity of the repair, such as housing leak repair should arrive on site for repair within 1 hour, wall detachment repair should be repaired within 3 working days; public area cleaning complaints should arrange cleaning work within 1 hour. For enterprise services, enterprise entry-related procedure consultation should provide detailed procedure handling guidelines and answers within 1 working day; service demand change applications should communicate with the enterprise and feedback processing opinions within 3 working days; enterprise activity organization assistance should start interfacing specific matters with the enterprise 5 working days before the activity to ensure smooth activity progress.

[0097] Flow path rules: initial processing path, service requests are assigned to the corresponding service team and staff through the above task allocation rules, entering the initial processing stage. Staff need to respond to service requests within the processing time limit, confirm receipt of the task and start processing. Staff need to take appropriate measures according to the specific content of the service request, such as on-site investigation, troubleshooting, data query, communication and coordination, etc., and record the processing process and progress in the system. Internal flow path, during processing, if it is found that the cooperation of other departments is needed, the staff can initiate an internal flow application in the system to transfer the task to the relevant support department or team. For example, during equipment maintenance, it is found that the design of the equipment needs to be changed to reform the energy system, the maintenance personnel initiate an internal flow application in the system to transfer the task to the energy management department for collaborative processing. After receiving the transferred task, the support department or team needs to continue processing according to the specified processing time limit and feed back the processing results to the original staff. The original staff continue to promote the task according to the feedback results to ensure that the service request is solved as a whole. External collaboration path, for some service requests that need the support of external enterprises in the park, such as the procurement of special materials, the system supports the transfer of tasks to external collaboration units. Staff send detailed collaboration demand information to external collaboration units through the system and track the processing progress. After the external collaboration unit completes the task, it feeds back the processing results to the system, and the staff audit and confirm the processing results. If it meets the requirements, mark the task as completed; if it does not meet the requirements, communicate with the external collaboration unit to determine further solutions and continue to track the processing progress until the task is completed.

[0098] Service execution: when the user inputs the service request through the given channel and submits, the system uses NLP technology to analyze the user's input service request, identifies the service type, then matches the predefined processing flow, assigns the service task to the corresponding processing node, and synchronizes the processing progress data to the fusion database in real time.

[0099] Service feedback and optimization: The system uses Apache Kafka and Apache to collect the allocation status, processing progress, and time limit data of real-time tasks, and then uses the LSTM model to predict process deviations and feedback to users and administrators in a timely manner. After the service is completed, the staff needs to fill in the detailed service request resolution, measures taken, and time nodes in the system, and submit them to the user for feedback. At the same time, the system triggers the email / sms API through the RabbitMQ information queue, and the user receives the service evaluation questionnaire to express their satisfaction and opinions. The user evaluation text is analyzed by the BERT model, such as "waited for too long" as negative emotions. Based on the DQN algorithm, the analysis results are dynamically decided, simple problems are automatically returned to the original staff, complex problems are upgraded to supervisors, and the service process closed-loop management is realized until the user is satisfied. The system monthly analyzes and analyzes various data in the service request processing process, such as the number trend of different types of service requests, the average processing time, the transfer efficiency of each link, and the user satisfaction evaluation distribution, and automatically outputs the monthly analysis report using Python and Matplotlib, and generates optimization suggestions based on NLP technology to optimize and adjust the service request processing process. For example, optimize the task allocation rules to improve the accuracy and rationality of task allocation; adjust the processing time limit to better meet user needs and actual work conditions; simplify the transfer link to reduce unnecessary time cost. Regularly evaluate the effect of process optimization measures, such as comparing key indicator data before and after optimization, including: processing time limit, customer satisfaction, to verify the effectiveness of optimization measures. If the expected optimization effect is not achieved, analyze the reasons and further adjust the optimization strategy to continuously improve the service request processing process and improve the overall performance and user experience of the intelligent service model.

[0100] The service feedback and optimization also includes handling of service overtime and rule conflict situations. When service overtime occurs, an early warning mechanism is automatically triggered and the task is upgraded to the superior management department or complaint handling department; for overtime caused by special reasons, the staff needs to submit an application containing the reason for the delay and the expected completion time, and adjust the processing time limit after approval. At the same time, the system will push the overtime explanation and solution to the user in real time, and all processing state changes are synchronized to the fusion database in a timely manner. For rule conflict situations, the system automatically freezes the current process and transfers it to the manual review node to ensure that the entire abnormal disposal is traceable.

[0101] Data security and privacy protection module: used for data encryption and desensitization, access control, and real-time monitoring and recording of data operation behavior to protect data security and user privacy;

[0102] It should be noted that the AES encryption algorithm is used to encrypt the sensitive data stored in the cloud data center. In the process of data transmission, the SSL / TLS protocol is used to establish an encrypted channel. The RBAC model is used to assign different roles according to the responsibilities and permissions of the park users, and each role has corresponding data access permissions. Various identity authentication technologies such as dynamic password and biometric recognition are used to ensure the authenticity of the user's identity. When analyzing service requests, determine the data fields that need to be desensitized and the degree of desensitization, such as using partial shielding for the name field and replacing the ID number field for desensitization. In the process of data mining and analysis, data sharing, etc., the desensitized data is applied to ensure the privacy and security of user data. Deploy a data leakage prevention system to monitor data access, modification, transmission and other operation behaviors in real time; set security policies and alarm thresholds, and when abnormal data access behavior is detected, timely warning and corresponding measures are taken to intercept and handle. Establish a data audit system to record detailed information of all data operations, including operation user, operation time, operation type, operation object, etc. Regularly analyze and review the audit logs to timely discover potential security problems and risks.

[0103] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and variations can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A smart park collaborative service system based on cloud computing and big data, including a data security and privacy protection module, characterized in that: Data acquisition module: used for real-time acquisition of multi-source data in the smart park; Data processing module: used to build data fusion and collaboration models, extract and fuse features from multi-source data in smart parks, and then perform collaborative analysis based on big data to generate collaborative decision-making suggestions and achieve data interoperability and sharing; Intelligent Service and Optimization Module: Used to build intelligent service models, automatically generate services based on service requests, track and provide feedback on service progress in real time, optimize and update models, and improve service response efficiency.

2. The smart park collaborative service system based on cloud computing and big data according to claim 1, characterized in that: The method for creating the data fusion and collaboration model includes: H1: Data Processing: Standardize and clean the collected multi-source data of the smart park, and perform coordinate transformation and timestamp synchronization according to a unified spatiotemporal benchmark to obtain the processed multi-source data of the smart park. H2: Feature Fusion: Based on the spatiotemporal-semantic association knowledge network, the processed multi-source data of the smart park are cross-modal feature associations to establish a fusion database; H3: Decision Collaboration: Based on big data, analyze and integrate data from the database, explore the relationships between data, and generate collaborative decision suggestions.

3. The smart park collaborative service system based on cloud computing and big data according to claim 2, characterized in that: The method for constructing the spatiotemporal-semantic association knowledge network includes: N1: Data Labeling: Assigning preliminary semantic labels to the processed multi-source data of the smart park; N2: Geospatial Association Construction: Based on the processed multi-source data of the smart park, construct a spatial geographic model of the smart park; calculate the spatial relationships between different spatial elements, including: distance, adjacency, and containment; N3: Spatiotemporal element extraction: Analyze and process multi-source data of the smart park, extract time patterns, mine time patterns, and obtain spatiotemporal elements; N4: Semantic association knowledge extraction and fusion: Extract semantic knowledge of smart parks from multi-source data and relevant literature on smart parks after processing, and construct a semantic knowledge base; based on the semantic knowledge base, associate the multi-source data and semantic knowledge of smart parks after processing. N5: Network Construction: Using the processed multi-source data of the smart park, spatiotemporal elements, semantic knowledge, and spatial geographic models and their spatial relationships as nodes, and the association relationships between nodes and the weights of spatial relationships as edges, a spatiotemporal-semantic association knowledge network is constructed.

4. The smart park collaborative service system based on cloud computing and big data according to claim 2, characterized in that: The steps for generating collaborative decision-making recommendations include: P1: Data Analysis: Perform multi-dimensional analysis on the data in the integrated database. The multi-dimensional analysis includes: time dimension, spatial dimension, and modal dimension, to obtain the data analysis results; P2: Association rule mining: Based on the data analysis results, mine the association relationships between data in the integrated database; and predict future data trends and events based on the association relationships to obtain prediction results; P3: Generate decision-making suggestions: Construct decision-making scenarios, determine decision indicators and influencing factors; based on correlations and prediction results, generate collaborative decision-making suggestions for the corresponding scenarios.

5. A smart park collaborative service system based on cloud computing and big data according to claim 2, characterized in that: The feature fusion method includes: M1: Cross-modal feature association: The processed smart park data is spliced ​​to generate a basic feature vector; multimodal features are weighted and aggregated to obtain a multimodal feature vector; M2: Dimensionality Reduction and Optimization: Dimensionality reduction is performed on the spatiotemporal-semantic association knowledge network and multimodal feature vectors to obtain dimensionality-reduced feature vectors. Feature weights are then adjusted to establish a fusion database.

6. The smart park collaborative service system based on cloud computing and big data according to claim 5, characterized in that: The method for cross-modal feature association includes: Q1: Concatenation: Connect the feature vectors of different modalities aligned within the same spatiotemporal unit to obtain the basic feature vector; Q2: Calculate feature weights: Automatically learn the specific content of the basic feature vectors, obtain the dynamic weight of each feature dimension, and highlight key information; Q3: Weighted aggregation: The calculated feature weights are summed with the basic feature vectors to obtain the multimodal feature vector.

7. A smart park collaborative service system based on cloud computing and big data as described in claim 5, characterized in that: The method for adjusting feature weights includes: E1: Normalized dimensionality reduction features: The dimensionality-reduced feature vectors are standardized, and the initial weights are redistributed based on the contribution of the feature variance. E2: Calculate dynamic weights: Analyze the semantic relationships between multimodal features and calculate dynamic weight coefficients; E3: Similarity-driven adjustment: Calculate the similarity matrix between feature vectors and adjust the feature weights according to the strength of the correlation.

8. The smart park collaborative service system based on cloud computing and big data according to claim 1, characterized in that: The method for creating the intelligent service model includes: S1: Service Function Integration: Integrating various service function modules; S2: Process Definition: Defines the processing flow for various services, including task allocation, processing time limits, and workflow paths; S3: Information Input: The user inputs a service request; S4: Service Execution: Based on the processing flow, determine the service type according to the service request, automatically generate a service, and transfer it to the relevant service; S5: Service Feedback and Optimization: Track the progress of service request processing in real time and provide feedback to users; after the service is completed, push a service evaluation survey to users, analyze the collected service evaluation data, and optimize the processing flow based on the analysis results.

9. A smart park collaborative service system based on cloud computing and big data according to claim 8, characterized in that: The service execution steps include: R1: Determine Service Type: Parse the user's service request, identify the service type, and match it with a predefined processing flow; R2: Service Generation: Based on the processing flow, service tasks are assigned to the corresponding processing nodes, and processing progress data is synchronized to the fusion database in real time.

10. A smart park collaborative service system based on cloud computing and big data according to claim 8, characterized in that: The service feedback and optimization includes handling service timeouts and rule conflicts. The handling methods include: automatically triggering an early warning mechanism and transferring the call to a manual review node, notifying the user of the reasons for the service timeout and the solution, and updating the fusion database.

Images