Campus shared equipment management system and method based on intelligent middle station

Abstract

The invention relates to the technical field of equipment operation and maintenance, and discloses a campus shared equipment management system and method based on an intelligent middle platform. The system comprises a plurality of shared devices and an intelligent middle station, and each shared device is provided with an identification module and a sensor group. The identification module displays basic data of the equipment, and the access module establishes a communication channel between the equipment and a middle station. The user evaluation module is used for calculating a user reputation score through historical data and a preset index; and the authority dynamic allocation module endows the user with an authority level according to the score. The prediction module inputs the operation data into the pre-training model to generate a demand prediction result; and the scheduling optimization module generates a scheduling scheme by adopting an intelligent optimization algorithm based on the multi-objective function and the authority level. The shared facility use process is simplified through reputation scoring, tedious procedures are not needed, and the management efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment operation and maintenance, and in particular to a campus shared equipment management system and method based on a smart middle platform. Background Art

[0002] With the popularization of the concept of smart campuses, shared equipment management systems have become key infrastructure for improving the utilization rate of campus resources. Currently, various types of shared equipment have gradually achieved basic Internet of Things access, enabling device search and basic reservation functions through mobile terminals. However, the rapid increase in the number of devices and the complexity of usage scenarios have exposed obvious bottlenecks in traditional management systems in aspects such as dynamic resource allocation, user behavior guidance, and device status maintenance, making it difficult to meet the high-concurrency and multi-scenario campus usage requirements.

[0003] In the prior art, mainstream shared equipment management systems mostly adopt the following technical routes: a device discovery mechanism based on location services to achieve device positioning through GPS or Bluetooth beacons; a single-dimensional credit evaluation method, such as freezing accounts based on device damage compensation records; a linear regression model based on historical reservation data to predict peak device usage periods; a first-come, first-served queuing algorithm for device allocation, freezing account permissions based on the number of historical user defaults, and adopting a time slice rotation strategy to allocate high-demand devices. Another typical solution generates a static device scheduling table by analyzing the periodic usage patterns of user groups.

[0004] However, the inventors found that the prior art has the following defects during the implementation of the present invention: (1) The credit evaluation mechanism only relies on records of extreme events such as device damage, lacking refined evaluation of users' daily usage behaviors, resulting in distorted credit portraits; (2) The demand prediction model does not consider real-time device status parameters (such as printer toner levels, fitness equipment wear), leading to a high prediction error rate in the event of sudden device failures; (3) The resource scheduling strategy is not sound, and key indicators such as device vacancy rates and user waiting times often deteriorate contradictorily; (4) The permission management adopts an "all-or-nothing" binary control mode, unable to implement differential service strategies. These problems seriously restrict the service quality and operational sustainability of the shared equipment system. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a campus shared equipment management system and method based on a smart middle platform to solve at least one of the above technical problems.

[0006] To achieve the above object, in a first aspect, a campus shared equipment management system based on a smart middle platform is provided, including:

[0007] A number of shared devices and a smart middle platform;

[0008] Each of the shared devices is provided with an identification module and a sensor group;

[0009] The sensor group is used to obtain the operation data of the shared device;

[0010] The identification module is used to display the basic data of the shared device to the user;

[0011] The intelligent middle platform includes an access module, a user evaluation module, a permission dynamic allocation module, a prediction module, and a scheduling optimization module;

[0012] The access module is used to establish a communication channel between the shared device and the intelligent middle platform;

[0013] The user evaluation module is used to obtain historical data, and obtain the user's credit score according to the historical data, preset user evaluation indicators, and a preset credit scoring model;

[0014] The permission dynamic allocation module is used to assign a permission level to the user according to the credit score;

[0015] The prediction module is used to obtain the operation data collected by the sensor group, input the operation data into a pre-trained demand prediction model, and obtain a demand prediction result;

[0016] The scheduling optimization module is used to establish a multi-objective function according to the operation data, and then generate a scheduling plan for the demand prediction result by using an intelligent optimization algorithm according to the multi-objective function and the permission level.

[0017] In a second aspect, the present invention provides a method for managing campus shared devices based on an intelligent middle platform. Based on an intelligent middle platform-based campus shared device management system in the first aspect, the method includes the following steps:

[0018] Establish a communication channel between the shared device and the intelligent middle platform;

[0019] Historical data, and obtain the user's credit score according to the historical data, preset user evaluation indicators, and a preset credit scoring model;

[0020] Assign a permission level to the user according to the credit score;

[0021] Obtain the operation data of the shared device from the sensor group, input the operation data into a pre-trained demand prediction model, and obtain a demand prediction result;

[0022] Establish a multi-objective function according to the operation data, and then generate a scheduling plan for the demand prediction result by using an intelligent optimization algorithm according to the multi-objective function and the permission level.

[0023] In a third aspect, an electronic device is provided, which includes:

[0024] One or more processors;

[0025] A storage device for storing one or more programs,

[0026] When the one or more programs are executed by the one or more processors, the one or more processors implement a campus shared device management method based on an intelligent middle platform as described in the second aspect.

[0027] In a fourth aspect, there is provided a computer-readable storage medium having stored thereon a computer program, which when executed by a processor implements a campus shared device management method based on an intelligent middle platform as described in the second aspect.

[0028] The above technical solution has the following beneficial technical effects:

[0029] In the present invention, by calculating the user credit score and assigning a user permission level according to the credit score, without cumbersome procedures, the usage process of shared facilities is simplified. Then, a demand prediction result is obtained through operation data, and a scheduling plan is obtained by optimizing the demand prediction result through establishing a multi-objective function, thereby improving the flexibility of shared device management.

[0030] By constructing a multi-dimensional collaborative management architecture of an intelligent middle platform, the intelligent linkage management of user behavior and device resources is realized. Based on the combination of the device operation status data collected in real time by the sensor group and the pre-trained demand prediction model, the limitations of traditional single historical data analysis are avoided, the accuracy and timeliness of shared device demand prediction are improved, and sudden usage scenarios are effectively responded to. The user evaluation module constructs a fine-grained credit evaluation system by integrating multi-dimensional historical behavior data and dynamically updated evaluation indicators, realizes the accurate mapping of user permission levels, while improving the service priority of high-credit users, strengthens the restraint mechanism for abnormal usage behaviors. The scheduling optimization module adopts a collaborative decision-making mechanism of multi-objective functions and intelligent optimization algorithms, takes into account multiple optimization objectives such as device resource utilization rate, maintenance cost, and user waiting time, and dynamically adjusts the resource allocation strategy in combination with the permission level to achieve the balance of service fairness and system operation efficiency. The coordinated operation of the identification module and the access module ensures the real-time synchronization of device status information and user terminals, improves the user operation response efficiency and device management transparency. In short, through modular function collaboration and data closed-loop circulation, the service stability, resource scheduling rationality, and management intelligence level of the campus shared device system are comprehensively improved. Description of the Drawings

[0031] The drawings are used to better understand the present invention and do not constitute an improper limitation to the present invention. Among them:

[0032] Figure 1It is a structural block diagram of a campus shared device management system based on a smart middle platform in an embodiment of the present invention;

[0033] Figure 2 It is a structural block diagram of the recognition module in an embodiment of the present invention;

[0034] Figure 3 It is a structural block diagram of the access module in an embodiment of the present invention;

[0035] Figure 4 It is a structural block diagram of the user evaluation module in an embodiment of the present invention;

[0036] Figure 5 It is a structural block diagram of the permission dynamic allocation module in an embodiment of the present invention;

[0037] Figure 6 It is a structural block diagram of the prediction unit in an embodiment of the present invention;

[0038] Figure 7 It is a structural block diagram of the optimization scheduling module in an embodiment of the present invention;

[0039] Figure 8 It is a flowchart of a campus shared device management method based on a smart middle platform in an embodiment of the present invention;

[0040] Figure 9 It is a structural schematic diagram of a computer system in an embodiment of the present invention. Specific Embodiments

[0041] The following describes exemplary embodiments of the present invention with reference to the accompanying drawings, including various details of the embodiments of the present invention to facilitate understanding, which should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and conciseness, the description below omits the description of well-known functions and structures.

[0042] Embodiment 1

[0043] As Figure 1 shown, this embodiment provides a campus shared device management system based on a smart middle platform, including:

[0044] A number of shared devices and a smart middle platform;

[0045] Each of the shared devices is provided with a recognition module and a sensor group;

[0046] The sensor group is used to obtain the operation data of the shared device;

[0047] The recognition module is used to display the basic data of the shared device to the user;

[0048] The intelligent middleware platform includes an access module, a user evaluation module, a dynamic permission allocation module, a prediction module, and a scheduling optimization module;

[0049] The access module is used to establish a communication channel between the shared device and the intelligent middleware platform;

[0050] The user evaluation module is used to obtain historical data, and obtain the user's credit score according to the historical data, preset user evaluation indicators, and a preset credit scoring model;

[0051] The dynamic permission allocation module is used to assign a permission level to the user according to the credit score;

[0052] The prediction module is used to obtain the operation data collected by the sensor group, input the operation data into a pre-trained demand prediction model, and obtain a demand prediction result;

[0053] The scheduling optimization module is used to establish a multi-objective function according to the operation data, and then generate a scheduling plan for the demand prediction result by using an intelligent optimization algorithm according to the multi-objective function and the permission level.

[0054] Specifically, as Figure 2 shown, in the recognition module, it specifically includes:

[0055] The basic data acquisition unit is used to acquire the basic data of the shared device;

[0056] The basic feature extraction unit is used to extract features from the basic data to obtain a number of basic feature vectors;

[0057] The digital identifier generation unit is used to generate a digital identifier by using a hash algorithm according to a number of the basic feature vectors;

[0058] The identification code generation unit is used to generate an identification code according to the basic data and the digital identifier, and the identification code is set on the surface of the shared device.

[0059] Specifically, the basic data includes the device model, technical specifications, factory date, expected service life, etc. of the shared device.

[0060] Specifically, the basic feature vectors include the device serial number, model parameters, factory batch, etc. The basic feature vector V is represented as a multi-dimensional feature matrix, which contains the static feature information and dynamic feature information of the shared device. For example, for a shared printer, its basic feature vector V contains the following dimensions:

[0061] V = [device type code, manufacturer ID, factory date, serial number, hardware configuration parameters, technical specifications, initial use status];

[0062] In the formula, the device type code is, for example, "PRINTER_001"; the manufacturer ID is, for example, "HP_MANUFACTURER_CODE"; the factory date format is "YYYYMMDD", for example, "20250205"; the serial number is, for example, "SN2025061501"; the hardware configuration parameters include memory size, processor type, etc.; the technical specifications include printing speed, resolution, consumable type, etc., and the initial use state is a new product status mark. Such a basic feature vector V not only contains static information but also can reflect the basic attributes and initial state of the device.

[0063] Specifically, in the digital recognition unit, the calculation formula of the digital identifier is as follows:

[0064] DID = Hash(concat(V, T, R));

[0065] In the formula, DID is the digital identifier, Hash() is an encryption hash function (such as SHA-256), T is the timestamp, R is the random salt value, and concat() is a normalized information concatenation function. The concat() is used to securely and consistently concatenate data elements of different types and formats according to predefined rules. This function first performs format unification and type conversion on the input data to ensure that even data from different sources and with different structures can be reliably combined. It is not just simple string concatenation but an advanced concatenation operation with data preprocessing, type normalization, and security conversion capabilities. In the scenario of digital identifier generation, concat() can concatenate the device feature vector, timestamp, and random salt value in a fixed order and with a specific delimiter to provide a standardized input for subsequent hash encryption, thereby enhancing the uniqueness and irreversibility of the generated digital identifier. The digital identifier is an immutable digital archive, and each digital archive includes detailed information sharing the entire life cycle of the device, and blockchain technology is used to ensure the security and traceability of the data. Each device status change leaves an immutable record on the blockchain, forming a complete device digital archive.

[0066] Specifically, the identification code is a high-precision Radio Frequency Identification (RFID) electronic tag, an intelligent barcode, or a two-dimensional code. Users can scan the identification code through an intelligent terminal (such as a mobile phone or a tablet computer, etc.) to read out the digital identifier corresponding to the shared device, and obtain the basic information of the shared device according to the digital identifier. Taking a shared laboratory instrument as an example, on an electron microscope, a high-frequency RFID electronic tag and an embedded two-dimensional code are installed first. When the user approaches the instrument with an intelligent terminal, all the basic information of the electron microscope can be instantly read, such as the device model (Zeiss AxioObserver Z1), the factory date (March 2022), the asset number (LAB-MI-2022-003), etc. At the same time, the sensor group continuously collects the key operating parameters of the instrument, such as the lens temperature, the motor speed, the alignment accuracy of the optical system, and the cumulative working hours of use, and transmits them to the intelligent middle platform in real time through the wireless network.

[0067] Specifically, as Figure 3 shown, in the access module, it specifically includes:

[0068] A protocol setting unit for initiating a connection request with the intelligent middle platform through a preset standard access protocol in the shared device;

[0069] A hardware fingerprint calculation unit for calculating the hardware fingerprint hash value of the shared device according to the basic data;

[0070] A network identification code allocation unit for generating a network identification code according to the hardware fingerprint hash value and a number of the basic feature vectors;

[0071] A communication unit for maintaining communication between the intelligent middle platform and the shared device according to the network identification code.

[0072] Specifically, in the protocol setting unit, the preset standard access protocol can be set within the sensor group or a separate signal module can be set on the shared device. The standard access protocol can adopt IEEE802.1X. The standard access protocol initiates the connection process and sends a connection request to the intelligent middle platform while carrying its basic feature vector V. The intelligent middle platform will verify the basic communication parameters of the shared device, such as the supported network protocol type, encryption standard, and authentication method, etc. For example, a printer uses the Extensible Authentication Protocol - Transport Layer Security (EAP-TLS) as its access protocol. The intelligent middle platform, as the authentication server, fully executes the device verification process according to the IEEE802.1X standard. When a shared printer sends a network access request while carrying its basic feature vector V, the campus intelligent middle platform will immediately initiate multi-dimensional security checks: First, verify the device identity information and credentials through EAP-TLS, then check the supported network protocol type, encryption standard, and authentication method, and comprehensively evaluate the security and compatibility of the shared device. Based on the preset access policy, the campus intelligent middle platform will determine whether the device has the permission to enter a specific network area, allocate corresponding network resources and access permissions to the devices that pass the verification, while for the devices that fail to meet the security requirements, directly reject the access and record detailed exception logs to ensure the overall security and controllability of the campus network environment.

[0073] Specifically, in the hardware fingerprint calculation unit, the hardware fingerprint hash value is a security authentication mechanism based on the physical characteristics of the device. By capturing the unique hardware characteristics of the shared device, such as the MAC address, processor ID, firmware characteristics, etc., an irreproducible hardware fingerprint is generated. When capturing the unique hardware characteristics of the shared device, it is carried out according to the basic data. The calculation formula of the hardware fingerprint hash value is as follows:

[0074] H = Hash(M||P||F);

[0075] Wherein, H is the hardware fingerprint hash value, Hash() is an encryption hash function (such as SHA-256), and "||" represents string concatenation. M is the unique identification code of the device network interface (6 bytes in length), P is the processor identification code (a 12-16-bit random sequence), and F is the firmware version code (a combination of 8 digits and letters). M represents the physical address of the device network interface and is the globally unique identification code of the network card or communication module. P reflects the inherent characteristics of the processor, similar to the fingerprint in biometrics, and records the micro-structure and manufacturing characteristics of the processor. F records the version information of the device firmware and contains the uniqueness mark at the software level. The above concatenation order helps to ensure that the generated hash value has the maximum entropy and randomness. This step ensures that only verified devices can access the intelligent middle platform.

[0076] Specifically, in the network identification code allocation unit, the calculation formula of the network identification code is as follows:

[0077] NetID = Hash(V'||H||T');

[0078] Wherein, NetID is the network identification code; H() is the hash function; V' is the string corresponding to the basic feature vector; H is the hardware fingerprint hash value; T is the time stamp, that is, the exact time stamp when the unique network identification code is generated, which is used to ensure that each generated network identification code has a time dimension and prevent replay attacks. When calculating the network identification code, first convert the basic feature vector V into a JSON string to obtain V', and then concatenate V' with the hardware fingerprint H and the time stamp T in sequence. Use a strong encryption hash function such as SHA-256 to hash the concatenated string to generate the final network identification code. This process ensures that the generated identification code has high randomness, uniqueness, and unpredictability.

[0079] Specifically, after generating the network identification code, the shared device and the intelligent middle platform establish a real-time communication channel based on protocols such as WebSocket or MQTT. This channel supports two-way and low-latency data exchange, enabling the device to continuously report its status, receive configuration instructions, and support remote management and monitoring.

[0080] Specifically, in the access module, there is also an encryption unit, which is arranged to execute after the hardware fingerprint calculation unit. The encryption unit is used to establish an encrypted communication channel between the shared device and the intelligent middle platform. By adopting the Public Key Infrastructure (PKI) and using an asymmetric encryption algorithm (for example, RSA (Rivest-Shamir-Adleman) or Elliptic Curve Cryptography (ECC)), a symmetric session key is negotiated. The process of establishing the encrypted communication channel is as follows: the shared device encrypts the temporary session key with the public key of the intelligent middle platform, and the intelligent middle platform decrypts and verifies the session key with the private key. Both parties use the negotiated session key for subsequent encrypted communication.

[0081] Specifically, as Figure 4 shown, in the user evaluation module, it specifically includes:

[0082] A historical data acquisition unit, which is used to acquire historical data;

[0083] An index determination unit, which is used to obtain the values of several user evaluation indexes according to the preset user evaluation indexes and the historical data;

[0084] A reputation score calculation unit, which is used to input the values of several user evaluation indexes into a preset reputation score model to obtain the user's reputation score.

[0085] Specifically, the historical data is the historical data of user equipment usage, which includes whether equipment usage training has been carried out, the number of times of illegal use, the history of equipment damage, and the degree of compliance with usage specifications, etc.

[0086] Specifically, the preset user evaluation indexes include equipment usage behavior indexes, integrity-related indexes, equipment operation proficiency indexes, and environmental protection performance indexes.

[0087] The equipment usage behavior index is an important index for evaluating the user's reputation. This includes whether the operations during equipment usage comply with the regulations, whether the equipment intact rate is maintained well, and whether it is cleaned and returned in time after use. By tracking the entire process of each equipment usage by the user, recording whether their operations are standardized and whether there are improper usage behaviors, and accordingly establishing a user's equipment usage reputation file. Through intelligent sensors and the background management system, the details of the user's equipment usage can be accurately captured to form an objective and fair evaluation system.

[0088] The integrity-related indicators are the basis for shared device management. The system will focus on key credibility indicators such as the number of times a user violates regulations, the record of compensating for damaged devices, and whether there are false repair reports. These data not only reflect the direct behavior of users, but also embody their respect and sense of responsibility for public resources. By establishing a comprehensive credit tracking mechanism, the system can accurately identify and record users' integrity performance.

[0089] The device operation proficiency indicators directly affect the use efficiency and safety of the device. The system will evaluate users' device operation proficiency, whether they have completed relevant usage training, and the device usage safety record, etc. These indicators can be comprehensively evaluated through various methods such as practical operation tests, training certifications, and usage logs. For complex or precision devices, the system can set a more stringent technical ability access mechanism to ensure the safe and efficient use of the devices.

[0090] The environmental performance indicators track indicators such as users' energy-saving usage records, resource utilization efficiency, and sustainable usage behaviors. This not only reflects users' environmental awareness, but also conforms to the green development concept of modern campuses. By setting corresponding incentive mechanisms, users can be guided to form more economical and environmentally friendly device usage habits.

[0091] Specifically, when calculating the values of several of the user evaluation indicators, first perform feature extraction on the historical data according to the user evaluation indicators. Set several sub-indicators under each user evaluation indicator, then assign values to each sub-indicator according to the historical data, and finally perform weighted summation according to the values of each sub-indicator to obtain the values of each user evaluation indicator.

[0092] Specifically, in the credit score calculation unit, the expression of the preset credit score model is as follows:

[0093] R = f(Σ(Wi * Xi));

[0094] In the formula, R represents the credit score, Wi represents the weight of the i-th user evaluation indicator, Xi represents the value of the i-th user evaluation indicator. The user evaluation indicators are not limited to the above four categories and can be subdivided into device usage compliance scores, historical violation records, device integrity, usage frequency and efficiency, and safety compliance scores. The safety compliance score, this indicator can be more specifically quantified, including: whether the system safety specifications are complied with, whether data protection is carried out as required, whether actively cooperate with safety inspections, whether there is a risk of sensitive information leakage, and whether the system and software are updated as required.

[0095] Specifically, as Figure 5 shown, in the dynamic permission allocation module, it specifically includes:

[0096] The user daily evaluation unit is used to obtain users' daily data;

[0097] A dynamic credit score adjustment unit for inputting the credit score and the daily data into a pre-trained machine learning model to obtain a dynamic credit score;

[0098] A permission allocation unit for allocating a corresponding permission level to a user according to the dynamic credit score and a preset permission grading.

[0099] Specifically, in the user daily evaluation unit, the daily data is obtained through a user behavior monitoring unit, which is a multi-level and multi-dimensional data collection and analysis platform. It captures the digital footprints of users in real time through intelligent sensors and log recording modules deployed in network devices, academic platforms, and campus systems. These data include, but are not limited to, device usage duration, accessed resource types, etc.

[0100] Specifically, in the dynamic credit score adjustment unit, the daily data is first preprocessed, and the preprocessing includes normalization, denoising, and feature engineering; then feature extraction is performed on the daily data to convert the multi-dimensional daily data into a format recognizable by machine learning, obtaining several daily features; the machine learning model is trained using the historical data, and by training the machine learning model with the historical data, the changes and trends of user behavior are identified; finally, the credit score is input into the trained machine learning model to obtain a dynamic credit score.

[0101] Specifically, in the permission allocation unit, users are divided into five permission levels: junior users (0 - 20 points), basic users (20 - 40 points), intermediate users (40 - 60 points), senior users (60 - 80 points), and privileged users (80 - 100 points). Each permission level corresponds to different device usage permissions, resource access scopes, and customized services, realizing differentiated and precise management of permission allocation. Based on the performance of users at each permission level, incentive strategies and penalty strategies for credit scores are designed. Positive behavior of users is given a credit score increase, and negative behavior is correspondingly given a credit score decrease. Through this mechanism, users are guided to maintain responsible device usage behavior and continuously improve their own permission levels. According to the finally determined permission level of the user, differentiated services and resource access permissions are provided. Junior users obtain basic functions and limited resources, while privileged users enjoy the highest-level customized services and extended resources, motivating users to continuously improve and upgrade.

[0102] Specifically, as Figure 6 shown, in the prediction module, it specifically includes:

[0103] A multi-dimensional data acquisition unit for acquiring multi-dimensional data of shared devices and establishing a multi-dimensional data set according to the multi-dimensional data;

[0104] A training unit for dividing the multi-dimensional dataset into a training set, a validation set, and a test set, and training a demand prediction model according to the training set;

[0105] A validation unit for validating the demand prediction model according to the validation set. If the validation is passed, the prediction unit is executed. If the validation fails, the training unit is returned to adjust the parameters of the demand prediction model and train a new demand prediction model;

[0106] A prediction unit for inputting the test set into the demand prediction model to obtain a demand prediction result.

[0107] Specifically, the multi-dimensional data includes operation data, enterprise management system data, remote monitoring data, and equipment operation and maintenance record data, and also includes data such as operation status, usage frequency, geographical location, energy consumption, performance indicators, and task response. The collected multi-dimensional raw data needs to be standardized and data-cleaned to remove noise and outliers in the multi-dimensional raw data, handle missing values, and verify the consistency of the multi-dimensional raw data, so as to ensure the quality and integrity of the data. Through systematic data collection and preprocessing, a high-quality and multi-dimensional equipment usage dataset is constructed.

[0108] Specifically, the multi-dimensional training set is divided into a training set, a validation set, and a test set according to a ratio of 8:1:1, and the division ratio can also be modified according to actual needs. The demand prediction model can be trained using a Long Short-Term Memory (LSTM) network or a random forest model. When training the demand prediction model, feature extraction is performed on the training set to obtain a number of training features, which include equipment type, usage frequency, time period, or user attributes, etc. Then, a number of the training features are input into the demand prediction model to be trained for training. The cross-validation method is used to optimize the hyperparameters of the model, and the best parameter combination is found through grid search.

[0109] Specifically, in the validation unit, the validation set is input into the trained demand prediction model to obtain a number of validation results. Then, metrics such as root mean square error and mean absolute error are calculated according to the number of validation results. If they all meet the preset range, the prediction unit is executed. If the validation fails, the training unit is returned to adjust the parameters of the demand prediction model and train a new demand prediction model.

[0110] Specifically, the prediction result not only outputs a point prediction value but also provides a confidence interval, providing data-driven support for resource scheduling decisions. Through continuous model monitoring and iterative optimization, the prediction accuracy and adaptability are continuously improved.

[0111] Specifically, as Figure 7 shown, in the optimization scheduling module, it specifically includes:

[0112] A constraint unit, configured to obtain the operation data and establish a multi-objective function according to the permission level and the operation data;

[0113] An optimization unit, configured to optimize the demand prediction result by using an intelligent optimization algorithm according to the multi-objective function to obtain an optimized demand result;

[0114] An output unit, configured to generate and output a scheduling scheme according to the optimized demand result and the operation data.

[0115] Specifically, in the constraint unit, the multi-objective function includes a permission level, equipment utilization rate, equipment response time, energy consumption cost, geographical distribution efficiency, etc. The equipment utilization rate is the ratio of the actual working time to the total available time. These values are set according to the operation data and the permission level.

[0116] Specifically, in the optimization unit, the intelligent optimization algorithm can be selected as a genetic algorithm or a particle swarm optimization algorithm, etc. Taking the particle swarm optimization algorithm as an example, it includes the following steps:

[0117] Randomly generate an initial particle swarm, where the initial particle swarm includes a number of initial particles, and each initial particle represents a scheduling scheme;

[0118] Call the multi-objective function;

[0119] According to the multi-objective function and using the Pareto domination principle, screen the initial particle swarm to obtain the Pareto front;

[0120] Update the initial particle swarm;

[0121] Repeat the above steps until a preset cut-off condition is met and output the final Pareto front;

[0122] According to the multi-objective function, screen the final Pareto front to obtain the optimized demand result.

[0123] Specifically, in the output unit, determine the status of each current shared device according to the real-time operation data, and then combine the optimized demand result to obtain a scheduling scheme. Under multi-dimensional constraint conditions, dynamically generate an optimal device allocation scheme to achieve real-time balance of system resources, maximize equipment utilization efficiency, reduce operation costs, and improve the overall system performance.

[0124] Specifically, the smart middle station also includes a maintenance module, which is used to monitor the health of shared equipment according to the operation data, and generate a maintenance plan to send to the maintenance personnel when a health hazard is detected. Based on the collected equipment operation data, a predictive maintenance model is constructed using machine learning algorithms (such as random forests, support vector machines, and neural networks). The predictive maintenance model can identify abnormal operation modes of equipment and predict potential failure risks. For example, by analyzing the historical maintenance records and real-time operation data of the printer, the model can predict when the print head needs to be replaced or which parts are about to reach the service life limit. According to the analysis results of the predictive maintenance model, the system automatically generates a personalized equipment maintenance plan. The equipment maintenance plan includes details such as maintenance time, replacement part type, and maintenance personnel arrangement. For example, for a frequently used shared printer, the system will arrange preventive maintenance during non-peak usage periods and prepare the required spare parts in advance. Then establish an intelligent maintenance resource scheduling mechanism to automatically match the most suitable maintenance personnel and resources based on factors such as equipment maintenance needs, maintenance personnel skills, and geographical location. The system can track the workload, professional skills, and geographical location of maintenance personnel in real time to ensure that maintenance work is performed efficiently and accurately. Each maintenance process is recorded in detail in an immutable file supported by blockchain technology. The records include maintenance time, maintenance personnel, replacement parts, maintenance costs, equipment status changes, etc. This method not only ensures the authenticity of maintenance records, but also provides a reliable data foundation for the full life cycle management of equipment.

[0125] Specifically, in the maintenance module, IoT sensors and edge computing technology are used to collect real-time equipment operation data, including key indicators such as temperature, vibration, current, and noise. For example, for a shared printer, the system can continuously monitor its internal temperature, print head working status, ink consumption, and mechanical parts wear. These sensors can collect multiple data points per second to build a real-time health portrait of the equipment's operation.

[0126] Taking shared printer monitoring as an example, in the intelligent monitoring of shared printers on campus, a comprehensive health status tracking system is built by deploying a sensor group. The temperature sensor monitors the internal temperature of the printer in real time to prevent overheating of key components; the current sensor tracks the current changes of the print head and paper feed mechanism to identify potential mechanical abnormalities; the optical sensor accurately detects the ink remaining and the degree of print head wear. The system collects 1-5 data points per second. When the temperature exceeds 65°C or the current fluctuates by more than 20%, the early warning mechanism is immediately triggered to ensure the safe and stable operation of the equipment.

[0127] Taking the monitoring of shared charging piles as an example, the monitoring system of campus shared charging piles adopts multi-sensor collaborative work. The voltage sensor monitors the output voltage stability in real time to ensure charging safety; the temperature sensor detects the temperature of the charging interface and internal components to prevent overheating risks; the current sensor tracks the current changes during the charging process to identify abnormal charging. The system collects 4-6 data points per second. When the voltage fluctuation exceeds ±5% or the interface temperature exceeds 70°C, an immediate warning will be issued to protect the charging equipment and user safety.

[0128] Taking the monitoring of shared bicycles as an example, for campus shared bicycles, this embodiment has developed a comprehensive health status monitoring system. The GPS positioning sensor tracks the vehicle location and usage trajectory; the vibration sensor detects the structural integrity of the frame and wheels; the electronic odometer records the riding mileage and usage frequency; the tire pressure sensor monitors the tire inflation pressure in real time. The system collects 1-2 data points per second. When the tire pressure is lower than 2.5 bar or abnormal vibration is detected, an alarm will be issued in a timely manner to ensure riding safety.

[0129] Specifically, in the process of constructing the predictive maintenance model, first, comprehensive equipment operation data is collected, including equipment operation parameters (temperature, vibration, current, rotation speed, etc.), historical maintenance records (maintenance time, maintenance type, replaced parts), sensor monitoring data (real-time performance indicators), usage environment data (humidity, dust, etc.) and equipment usage intensity data. In terms of algorithm selection, an ensemble learning method is adopted, comprehensively using random forest and neural network to give full play to their respective advantages. The specific model construction process includes: the data preprocessing stage, where multi-source heterogeneous data is cleaned, standardized and feature engineered to extract key features reflecting the equipment health status; the model training stage, where the dataset is divided into a training set and a test set, the random forest algorithm is used to learn the non-linear complex relationship of equipment failures, and the neural network (such as LSTM) captures the potential patterns in the time series; the model evaluation uses metrics such as precision, recall, and F1 score, and cross-validation is used to ensure the generalization performance of the model. The model can ultimately achieve early warning of equipment failures, not only predicting the probability of failure occurrence, but also estimating the remaining service life, providing data support for accurate and proactive maintenance decisions, and significantly reducing the risk of unplanned downtime and maintenance costs.

[0130] Specifically, the innovation of the intelligent scheduling mechanism for maintaining resource coordination is mainly reflected in the construction of a multi-dimensional, dynamic, and intelligent resource matching model. Traditional methods usually allocate resources based on static rules and manual experience, while this embodiment proposes an intelligent matching algorithm based on multi-objective optimization. The innovation in establishing the resource matching model introduces a comprehensive scoring mechanism and constructs a matching function that includes multiple dimensions of professional skills, geographical location, workload, and availability: M = f(professional skills, geographical location, workload, availability multi-dimensions). Compared with the traditional model, this model not only considers skill matching but also dynamically balances the workload of personnel to achieve more accurate resource allocation.

[0131] Specifically, the matching algorithm design adopts a hybrid method of integer programming and the Hungarian algorithm, extending the constraint conditions to workload balance and the minimum skill matching degree requirement. It not only pursues the optimal cost but also ensures the precise matching of maintenance tasks and the maximum efficiency of resources.

[0132] Specifically, by collecting maintenance process data in real time, the weight parameters of the matching algorithm are continuously optimized. This dynamic learning ability enables the resource scheduling model to continuously improve itself and intelligently adjust the matching strategy according to the actual operating conditions. Compared with traditional manual scheduling or simple rule matching, this solution improves the accuracy and efficiency of maintenance resource allocation and provides a more intelligent solution for enterprise equipment maintenance management.

[0133] Specifically, in the maintenance resource coordination model, each parameter has precise definitions and quantification standards. Skill (professional skills) is composed of the maintenance personnel's professional certificates, past maintenance experience, and proficiency in equipment types, and is quantified with a score of 0 - 100 to reflect the matching degree between the personnel and specific maintenance tasks. The weight of Location (geographical location) is based on the distance from the maintenance personnel to the equipment failure point, and the inverse distance weighted algorithm is used. The closer the distance, the higher the weight, comprehensively considering travel time and road conditions. The Workload coefficient calculates the number of tasks currently undertaken by the maintenance personnel, the complexity of the tasks, and the estimated completion time, etc., to ensure no over-allocation. The Availability (availability multi-dimensions) index includes multi-dimensional indicators such as the current working status of the personnel, rest time, and personal skill reserves, dynamically reflecting the ability of the personnel to be put into maintenance work in real time. Through the precise definition and correlation calculation of these parameters, a comprehensive, dynamic, and intelligent maintenance resource matching evaluation system is constructed.

[0134] Specifically, the intelligent middle platform further includes a space allocation module, which is used to obtain campus environment data, input the campus environment data into a pre-trained space matching model to obtain a scene matching score, and adjust the scheduling plan according to the scene matching score.

[0135] Specifically, in the space allocation module, the campus environment data can be obtained by using lidar, drone aerial photography, indoor positioning technology, etc., to draw detailed geographical information of campus buildings, teaching areas, and public spaces. For example, for a modern campus, the system can generate a three-dimensional space model accurate to the centimeter level, including building outlines, internal space layouts, corridors, classrooms, laboratories, etc. This model not only contains geometric information but also incorporates spatial attributes, such as room functions, usage intensities, and pedestrian flow densities.

[0136] Specifically, feature extraction is performed on the campus environment data to obtain a number of scenario features, which are used to analyze the usage patterns of devices at different times and in different spaces. The scenario features include usage frequency, usage time period, user group characteristics, functional requirements, etc. For example, for a shared printer, the system can identify its usage characteristics in different scenarios: mainly used for printing teaching materials in the teaching building, more for printing academic papers in the library, and for printing official documents in the administrative building.

[0137] Specifically, the space matching model is based on a machine learning model (e.g., a deep neural network or a support vector machine), and the expression of the space matching model is as follows:

[0138] M = Φ(δ, σ, υ);

[0139] In the formula, M represents the scene matching score, reflecting the matching degree between the device and the scene, Φ represents the non-linear mapping function, δ represents the device characteristic vector, σ represents the spatial attribute vector, and υ represents the usage requirement vector. The δ parameter includes quantitative characteristics such as the technical specifications, energy consumption level, and performance indicators of the device; the σ parameter covers the physical characteristics of the spatial environment, such as environmental factors like area, temperature, humidity, and illuminance; the υ parameter reflects the specific requirements of the user, including subjective and objective indicators such as function preferences, usage frequency, and expected performance. These three parameters are comprehensively evaluated through a complex non-linear mapping function, and finally, a matching score is output to quantify the fit between the device and a specific scene. f() is a complex non-linear mapping function implemented using a deep neural network or an ensemble learning model. Taking the printer matching in a shared office space as an example, this function comprehensively considers multi-dimensional input features. Assuming a Multilayer Perceptron (MLP) model, its input layer receives three types of parameters, δ, σ, and υ, and through non-linear activation functions (ReLU, Sigmoid) in multiple hidden layers, it performs feature interaction and weight learning. The model first standardizes each parameter, and then learns the complex non-linear mapping relationship between the parameters through fully connected layers. For example, when matching a printer to a specific office scene, the model will learn the association between "high-speed color printing requirements" and "open office space", and output a matching score between 0 and 1, reflecting the applicability of the device in this scene. This non-linear mapping can capture complex feature interaction patterns that are difficult for traditional linear models to identify.

[0140] Specifically, adjusting the scheduling plan according to the scene matching score can dynamically adjust the usage scenario of the device according to real-time data by establishing a device-scene adaptive algorithm. The system continuously monitors changes in space usage and fluctuations in user requirements, and automatically triggers the reallocation of devices. For example, during the exam week, some printing devices need to be temporarily allocated from the regular teaching areas to the exam preparation areas to meet the concentrated printing needs of students.

[0141] Specifically, taking the printer reallocation during the exam week as an example, during the final exam week, by monitoring the printing requirements in each teaching building, the system found that the printing volumes in the School of Humanities and the School of Science and Technology increased sharply, reaching 300% and 250% of the usual levels respectively. The algorithm automatically temporarily allocated some printing devices from the relatively idle administrative building and the Social Science Building to these two schools to relieve the device pressure during the printing peak period. At the same time, according to the size and urgency of the printing tasks, it gives priority to ensuring the rapid printing of exam-related documents.

[0142] Specifically, taking the dynamic adjustment of seats in the library's self-study area as an example, the library's intelligent system tracks the usage of seats in the self-study area in real time. Through cameras and seat sensors, it accurately captures the occupancy rate of seats in each area. When it is found that the seat utilization rate in the first-floor self-study area is as low as 30%, while the third-floor self-study area is full and there are a large number of students waiting, the system immediately activates the seat reallocation strategy. It guides students in the idle area to move orderly, or pushes real-time information about available seats through the display screen to balance the pressure of seat usage.

[0143] Specifically, taking the real-time optimization of campus network bandwidth as an example, during the period when students have classes intensively or there are large-scale online exams, the system monitors the network load. When it is found that the network bandwidth utilization rate in the teaching area reaches more than 85%, and the bandwidth utilization rate in some non-teaching areas (student apartments) is only 20%, the intelligent network management system will dynamically adjust the bandwidth allocation. It gives priority to ensuring the network quality in the teaching area and temporarily allocates additional bandwidth from other areas to ensure the network stability of online teaching and exams.

[0144] Specifically, taking the flexible allocation of classroom equipment (projectors, computers) as an example, in the multimedia teaching building, the system tracks the usage of equipment in each classroom. When it is found that the usage rate of projectors in some specialized course classrooms is low, while the equipment in general education classrooms is severely insufficient, it will automatically coordinate the equipment allocation. For example, the projector in the small seminar room of the School of Liberal Arts can be temporarily transferred to the large lecture hall to meet the equipment needs of different teaching scenarios and improve the utilization efficiency of teaching resources.

[0145] Specifically, taking the sharing and scheduling of laboratory instruments as an example, the cross-college laboratory instrument sharing platform monitors the usage of precision instruments in each laboratory in real time. When the usage rate of the electron microscope in the School of Biology is only 20%, while the demand for similar instruments in the School of Medicine is strong, the system will coordinate the temporary cross-college sharing of instruments. Through accurate usage scheduling and rapid physical transfer plans, it ensures the efficient utilization of valuable instruments, reduces duplicate purchases, and lowers the overall equipment investment cost of the school.

[0146] Specifically, the campus environment data provides the basic geographical and time information for equipment usage. Through accurate location coordinate tracking, the system can draw a spatial distribution heat map of equipment usage in each area of the campus. For example, it can accurately locate the equipment occupancy rate in different areas such as the library, teaching building, and laboratory, and analyze the usage intensity in each time period in combination with the time dimension. This kind of data can indicate the spatio-temporal rules of equipment usage and provide an intuitive visual basis for resource allocation.

[0147] Specifically, the above modules form a closed-loop intelligent decision-making system. First, real-time data is collected through multi-source sensors and data interfaces; then, key information is extracted from the original data using feature extraction algorithms; next, pattern recognition and trend prediction are performed through machine learning models; based on the recognition results, optimal resource allocation decisions are generated; finally, real-time reconfiguration of devices is achieved through an automated control system. This process is dynamic, adaptive, and intelligent, and can continuously optimize the utilization efficiency of campus device resources. A self-learning mechanism can also be added, where the system continuously collects device usage feedback and optimizes the scenario matching model. Each device usage will become training data for model improvement, making the scenario matching algorithm more and more accurate and better able to adapt to the dynamic changes in the campus environment.

[0148] Embodiment 2

[0149] As Figure 8 shown, this embodiment provides a campus shared device management method based on a smart middle platform, including the following steps:

[0150] S10: Establish a communication channel between the shared device and the smart middle platform;

[0151] S20: Obtain historical data, and obtain the user's credit score based on the historical data, preset user evaluation indicators, and a preset credit scoring model;

[0152] S30: Assign a permission level to the user according to the credit score;

[0153] S40: Obtain the operation data of the shared device from the sensor group, and input the operation data into a pre-trained demand prediction model to obtain a demand prediction result;

[0154] S50: Establish a multi-objective function according to the operation data, and then use an intelligent optimization algorithm to generate a scheduling plan for the demand prediction result according to the multi-objective function and the permission level.

[0155] Specifically, in the step S10, it includes the following steps:

[0156] S11: Initiate a connection request to the smart middle platform through a preset standard access protocol in the shared device;

[0157] S12: Calculate the hardware fingerprint hash value of the shared device according to the basic data;

[0158] S13: Generate a network identification code according to the hardware fingerprint hash value and several basic feature vectors;

[0159] S14: Maintain the communication between the smart middle platform and the shared device according to the network identification code.

[0160] Specifically, in the step S11, the preset standard access protocol can be set within the sensor group or a separate signal module can be set on the shared device. The standard access protocol can adopt IEEE802.1X. The standard access protocol initiates the connection process and sends a connection request to the intelligent middle platform carrying its basic feature vector V.

[0161] Specifically, in the step S12, the calculation formula for the hardware fingerprint hash value is as follows:

[0162] H = Hash(M||P||F);

[0163] In the formula, H is the hardware fingerprint hash value, Hash() is an encryption hash function (such as SHA - 256), "||" represents string concatenation. M is the unique identification code of the device network interface (6 - byte length), P is the processor identification code (a 12 - 16 - bit random sequence), and F is the firmware version encoding (an 8 - digit combination of numbers and letters).

[0164] Specifically, in the step S13, the calculation formula for the network identification code is as follows:

[0165] NetID = Hash(V'||H||T');

[0166] In the formula, NetID is the network identification code; H() is the hash function; V' is the string corresponding to the basic feature vector; H is the hardware fingerprint hash value; T is the timestamp, that is, the exact timestamp when generating the unique network identification code, which is used to ensure that each generated network identification code has a time dimension and prevent replay attacks. When calculating the network identification code, first convert the basic feature vector V into a JSON string to obtain V', and then concatenate V', the hardware fingerprint H, and the timestamp T in sequence. Use a strong encryption hash function such as SHA - 256 to hash the concatenated string to generate the final network identification code. This process ensures that the generated identification code has high randomness, uniqueness, and unpredictability.

[0167] Specifically, in the step S20, it includes the following steps:

[0168] S21: Obtain historical data;

[0169] S22: Obtain the values of several user evaluation indicators according to the preset user evaluation indicators and historical data;

[0170] S23: Input the values of several user evaluation indicators into the preset reputation scoring model to obtain the user's reputation score.

[0171] Specifically, the preset user evaluation indicators include device usage behavior indicators, integrity-related indicators, device operation proficiency indicators, and environmental protection performance indicators.

[0172] Specifically, in step S22, first, feature extraction is performed on the historical data according to the user evaluation indicators. A number of sub-indicators are set under each user evaluation indicator, and then the historical data is used to assign values to each sub-indicator. Finally, the values of each sub-indicator are weighted and summed to obtain the values of each user evaluation indicator.

[0173] Specifically, in step S23, the expression of the preset reputation scoring model is as follows:

[0174] R = f(Σ(Wi * Xi));

[0175] In the formula, R represents the reputation score, Wi represents the weight of the i-th user evaluation indicator, and Xi represents the value of the i-th user evaluation indicator. The user evaluation indicators are not limited to the above four categories and can be further divided into device usage compliance scores, historical violation records, device integrity levels, usage frequency and efficiency, and security compliance scores. The security compliance score can be more specifically quantified, including: whether the system security specifications are followed, whether data protection is carried out as required, whether security inspections are actively cooperated with, whether there is a risk of sensitive information leakage, and whether the system and software are updated as required.

[0176] Specifically, in step S30, the following steps are included:

[0177] S31: Obtain the daily data of the user;

[0178] S32: Input the reputation score and the daily data into a pre-trained machine learning model to obtain a dynamic reputation score;

[0179] S33: Assign the corresponding permission level to the user according to the dynamic reputation score and the preset permission grading.

[0180] Specifically, in step S31, the daily data is obtained through a user behavior monitoring unit, and the user behavior monitoring unit is a multi-level and multi-dimensional data collection and analysis platform.

[0181] Specifically, in step S32, first, preprocessing is performed on the daily data. The preprocessing includes standardization, denoising, and feature engineering; then feature extraction is performed on the daily data to convert the multi-dimensional daily data into a format recognizable by machine learning, obtaining a number of daily features; the historical data is used to train the machine learning model to identify changes and trends in user behavior through the historical data; finally, the reputation score is input into the trained machine learning model to obtain a dynamic reputation score.

[0182] Specifically, in the step S33, users are divided into five privilege levels: junior users (0 - 20 points), basic users (20 - 40 points), intermediate users (40 - 60 points), senior users (60 - 80 points), and privileged users (80 - 100 points).

[0183] Specifically, in the step S40, the following steps are included:

[0184] S41: Obtain multi - dimensional data of the shared device, and establish a multi - dimensional data set according to the multi - dimensional data;

[0185] S42: Divide the multi - dimensional data set into a training set, a validation set, and a test set, and train a demand prediction model according to the training set;

[0186] S43: Verify the demand prediction model according to the validation set. If it passes the verification, execute the prediction unit. If it fails to pass the verification, return to the training unit to adjust the parameters of the demand prediction model and train a new demand prediction model;

[0187] S44: Input the test set into the demand prediction model to obtain a demand prediction result.

[0188] Specifically, in the step S41, the multi - dimensional data includes operation data, enterprise management system data, remote monitoring data, and device operation and maintenance record data, and also includes data such as operation status, usage frequency, geographical location, energy consumption, performance indicators, and task response. The collected multi - dimensional raw data needs to be standardized and data - cleaned, removing noise and outliers in the multi - dimensional raw data, processing missing values, and verifying the consistency of the multi - dimensional raw data, so as to ensure the quality and integrity of the data. Through systematic data collection and pre - processing, a high - quality and multi - dimensional device usage data set is constructed.

[0189] Specifically, in the step S42, the multi - dimensional training set is divided into a training set, a validation set, and a test set according to the ratio of 8:1:1, and the division ratio can also be modified according to actual needs. The demand prediction model can be trained using a Long Short - Term Memory (LSTM) network or a random forest model. When training the demand prediction model, feature extraction is performed on the training set to obtain a number of training features, and the training features include device type, usage frequency, time period, or user attributes, etc. Then, a number of the training features are input into the demand prediction model to be trained. The cross - validation method is used to optimize the model hyperparameters, and the best parameter combination is found through grid search.

[0190] Specifically, in the step S43, the validation set is input into the trained demand prediction model to obtain several validation results, and then indicators such as root mean square error and mean absolute error are calculated based on the several validation results. If they all meet the preset range, the prediction unit is executed; if the validation fails, the training unit is returned to adjust the parameters of the demand prediction model and train a new demand prediction model.

[0191] Specifically, in the step S50, the following steps are included:

[0192] S51: Obtain the operation data and establish a multi-objective function based on the permission level and the operation data;

[0193] S52: Optimize the demand prediction result by using an intelligent optimization algorithm according to the multi-objective function to obtain an optimized demand result;

[0194] S53: Generate a scheduling plan based on the optimized demand result and the operation data and output it.

[0195] Specifically, in the step S51, the multi-objective function includes the permission level, equipment utilization rate, equipment response time, energy consumption cost, geographical distribution efficiency, etc.

[0196] Specifically, in the step S52, the following steps are included:

[0197] S521: Randomly generate an initial particle swarm, where the initial particle swarm includes several initial particles, and each initial particle represents a scheduling plan;

[0198] S522: Call the multi-objective function;

[0199] S523: Screen the initial particle swarm according to the Pareto dominance principle by using the multi-objective function to obtain the Pareto front;

[0200] S524: Update the initial particle swarm;

[0201] S525: Repeat steps S523 to S524 until the preset cut-off condition is met and output the final Pareto front;

[0202] S526: Screen the final Pareto front according to the multi-objective function to obtain the optimized demand result.

[0203] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiments can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction and do not limit the protection scope of this application. The specific working processes of the units and modules in the above system can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0204] An embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements any one of the above methods.

[0205] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, to implement all or part of the processes in the above method embodiments of the present invention, it can also be completed by a computer program instructing relevant hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal, and software distribution medium, etc. Of course, there are other ways of readable storage media, such as quantum memory, graphene memory, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0206] The present invention also provides an electronic device. The electronic device according to an embodiment of the present invention includes: one or more processors; a storage device for storing one or more programs, which, when executed by the one or more processors, enable the one or more processors to implement a campus shared device management method based on a wisdom middle platform provided by the present invention.

[0207] Reference is made below to Figure 9 , which shows a schematic structural diagram of a computer system 800 suitable for implementing the electronic device according to an embodiment of the present invention. Figure 9 The electronic device shown is merely an example and should not impose any limitations on the functions and scope of use of the embodiments of the present invention.

[0208] As Figure 9 shown, the computer system 800 includes a central processing unit (CPU) 801, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 802 or the program loaded from the storage section 808 into the random access memory (RAM) 803. In the RAM 803, various programs and data required for the operation of the computer system 800 are also stored. The CPU 801, ROM 802, and RAM 803 are connected to each other via a bus 804. The input / output (I / O) interface 805 is also connected to the bus 804.

[0209] The following components are connected to the I / O interface 805: an input section 806 including a keyboard, a mouse, etc.; an output section 807 including, for example, a cathode ray tube (CRT), a liquid crystal display (LCD), etc. and a speaker, etc.; a storage section 808 including a hard disk, etc.; and a communication section 809 including a network interface card such as a LAN card, a modem, etc. The communication section 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to the I / O interface 805 as required. A removable medium 811, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 810 as required so that a computer program read from it is installed into the storage section 808 as required.

[0210] Specifically, according to the embodiments disclosed by the present invention, the process described in the main step diagram above can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes program codes for performing the method shown in the main step diagram. In the above embodiment, the computer program can be downloaded and installed from the network through the communication section 809, and / or installed from the removable medium 811. When the computer program is executed by the central processing unit 801, the above functions defined in the system of the present invention are executed.

[0211] It should be noted that the computer-readable medium shown in the present invention can be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium can include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, the computer-readable storage medium can be any tangible medium that contains or stores a program, which can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present invention, the computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted by any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0212] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagram can represent a module, a program segment, or a part of code, and the above module, program segment, or part of code contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks can occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks shown can actually be executed substantially in parallel, and they can sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram or flowchart, and the combination of blocks in the block diagram or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0213] The units involved in the embodiments of the present invention can be implemented in software or in hardware. The described units can also be provided in a processor. For example, it can be described as: a processor includes a pre-response unit, a receiving unit, and a request unit.

[0214] The above specific embodiments do not limit the protection scope of the present invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A campus shared device management system based on an intelligent middle platform, characterized in that Including: A number of shared devices and an intelligent middle platform; Each of the shared devices is provided with an identification module and a sensor group; The sensor group is used to obtain the operation data of the shared device; The identification module is used to display the basic data of the shared device to the user; The intelligent middle platform includes an access module, a user evaluation module, a dynamic permission allocation module, a prediction module, and a scheduling optimization module; The access module is used to establish a communication channel between the shared device and the intelligent middle platform; The user evaluation module is used to obtain historical data, and obtain the user's credit score according to the historical data, the preset user evaluation index, and the preset credit scoring model; The dynamic permission allocation module is used to assign a permission level to the user according to the user's credit score; The prediction module is used to obtain the operation data collected by the sensor group, input the operation data into a pre-trained demand prediction model, and obtain a demand prediction result; The scheduling optimization module is used to establish a multi-objective function according to the operation data, and then generate a scheduling plan for the demand prediction result by using an intelligent optimization algorithm according to the multi-objective function and the permission level.

2. The campus shared device management system based on the intelligent middle platform according to claim 1, characterized in that, In the identification module, specifically including: A basic data acquisition unit, used to acquire the basic data of the shared device; A basic feature extraction unit, used to extract features according to the basic data to obtain a number of basic feature vectors; A digital identifier generation unit, used to generate a digital identifier by using a hash algorithm according to a number of the basic feature vectors; An identification code generation unit, used to generate an identification code according to the basic data and the digital identifier, and the identification code is set on the surface of the shared device.

3. The campus shared device management system based on the intelligent middle platform according to claim 2, wherein, In the access module, specifically including: A protocol setting unit, used to initiate a connection request with the intelligent middle platform through a preset standard access protocol in the shared device; A hardware fingerprint calculation unit, used to calculate the hardware fingerprint hash value of the shared device according to the basic data; A network identification code allocation unit, used to generate a network identification code according to the hardware fingerprint hash value and a number of the basic feature vectors; A communication unit, used to maintain the communication between the intelligent middle platform and the shared device according to the network identification code.

4. A campus shared device management system based on an intelligent middle platform according to claim 1, characterized in that, In the user evaluation module, specifically including: A historical data acquisition unit, used to acquire historical data; An index determination unit, used to obtain the values of a number of user evaluation indexes according to the preset user evaluation index and the historical data; A credit score calculation unit, used to input the values of a number of the user evaluation indexes into a preset credit scoring model to obtain the user's credit score.

5. A campus shared device management system based on an intelligent middle platform according to claim 1, characterized in that, In the dynamic permission allocation module, specifically including: A user daily data acquisition unit, used to acquire the daily data of the user; A dynamic credit score adjustment unit, used to input the credit score and the daily data into a pre-trained machine learning model to obtain a dynamic credit score; A permission allocation unit, used to allocate a corresponding permission level to the user according to the dynamic credit score and the preset permission classification.

6. The campus shared device management system based on the intelligent middle platform according to claim 1, characterized in that, In the prediction module, specifically including: A multi-dimensional data acquisition unit, used to acquire the multi-dimensional data of the shared device, and establish a multi-dimensional data set according to the multi-dimensional data; A training unit, configured to divide the multi-dimensional dataset into a training set, a validation set, and a test set, and train a demand prediction model according to the training set; A validation unit, configured to validate the demand prediction model according to the validation set. If the validation is passed, the prediction unit is executed. If the validation fails, the training unit is returned to adjust the parameters of the demand prediction model and train a new demand prediction model; A prediction unit, configured to input the test set into the demand prediction model to obtain a demand prediction result.

7. A campus shared equipment management system based on an intelligent middle platform according to claim 1, characterized in that, In the optimization scheduling module, it specifically includes: A constraint unit, configured to obtain the operation data and establish a multi-objective function according to the permission level and the operation data; An optimization unit, configured to optimize the demand prediction result by using an intelligent optimization algorithm according to the multi-objective function to obtain an optimized demand result; An output unit, configured to generate and output a scheduling plan according to the optimized demand result and the operation data.

8. A campus shared device management system based on an intelligent middle platform according to claim 1, characterized in that, The intelligent middle platform further includes a maintenance module, configured to perform health monitoring on shared devices according to the operation data, and generate a maintenance plan and send it to the repairman when a health hazard is detected.

9. A campus shared equipment management system based on an intelligent middle platform according to claim 1, characterized in that, The intelligent middle platform further includes a space allocation module, configured to obtain campus environment data, input the campus environment data into a pre-trained space matching model to obtain a scenario matching score, and adjust the scheduling plan according to the scenario matching score.

10. A campus shared device management method based on an intelligent middle platform, characterized in that, A campus shared device management system based on the intelligent middle platform according to any one of claims 1-9, the method includes the following steps: S10: Establish a communication channel between the shared device and the intelligent middle platform; S20: Obtain historical data, and obtain the user's credit score according to the historical data, a preset user evaluation index, and a preset credit scoring model; S30: Assign a permission level to the user according to the credit score; S40: Obtain the operation data of the shared device from the sensor group, input the operation data into a pre-trained demand prediction model to obtain a demand prediction result; S50: Establish a multi-objective function according to the operation data, and then use an intelligent optimization algorithm to generate a scheduling plan for the demand prediction result according to the multi-objective function and the permission level.

Images