A device state information active reporting method and system

Abstract

The application discloses a kind of device state information active reporting method and system, it is related to communication technical field, including collecting device historical data, predicting the state of device, and according to the prediction result dynamic adjustment reporting cycle, the data uploaded in each cycle is optimized, real-time monitoring network operating state, according to network state the data after optimization is uploaded to server.By collecting device historical data and predicting device state, dynamically adjust data sampling frequency and reporting cycle, optimize data collection efficiency, reduce the amount of invalid data transmission, improve resource utilization and reporting efficiency, the data in each cycle is compressed and priority division, ensure that high priority data is transmitted preferentially, reduce network transmission burden, using dynamic adjustment mechanism and intelligent algorithm, significantly improve the real-time performance and resource utilization efficiency of data upload.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a method and system for actively reporting device status information. Background Art

[0002] With the continuous development of information technology and the widespread adoption of the Internet of Things (IoT), smart devices, and various types of industrial equipment, device status monitoring and management have become crucial for ensuring efficient equipment operation and preventing failures. In network communications, in particular, the real-time collection and transmission of device status information is crucial for system stability and security. Existing device status monitoring systems mostly rely on periodic collection of device status information and upload it to a server via the network. However, these approaches often suffer from fixed sampling periods, inflexible transmission frequencies, and inadequate handling of network congestion. This leads to low information transmission efficiency, heavy network load, and even unnecessary data redundancy due to the dynamic nature of device status changes. While existing technologies have made some progress in the collection, upload, and prediction of device status information, they still face numerous shortcomings. Firstly, most existing methods rely on fixed sampling frequencies or periodic reporting, which cannot accurately match the dynamic nature of device status in systems. In particular, when device loads are excessive or temperatures are abnormal, device status changes frequently, and traditional methods cannot capture these changes in a timely manner. Secondly, existing network transmission solutions often focus on a single path selection or transmission protocol, lacking the ability to dynamically adjust transmission paths and protocols based on network conditions. Consequently, they cannot guarantee the real-time upload and reliability of device status information when network congestion or bandwidth is limited. In addition, data redundancy and network load management remain major challenges in the process of uploading device status information. Existing technologies lack effective data compression and optimization mechanisms, which often leads to inefficiency and delay in data transmission. Summary of the Invention

[0003] In view of the problems existing in the above-mentioned existing method and system for actively reporting device status information, the present invention is proposed.

[0004] Therefore, the present invention provides a method and system for actively reporting device status information to solve the shortcomings of the device status reporting method in terms of dynamism and intelligence, such as the inability to dynamically adjust the upload frequency, overly single path selection, low data processing efficiency, and lack of real-time perception and dynamic adjustment capabilities of the network status.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] In a first aspect, the present invention provides a method for actively reporting device status information, which includes collecting device historical data, predicting the device status, and dynamically adjusting the reporting period according to the prediction results;

[0007] Optimize the data uploaded in each cycle and monitor the network operation status in real time;

[0008] Upload the optimized data to the server according to the network status and store the device status information.

[0009] As a preferred solution of the method for actively reporting device status information described in the present invention, wherein: the collection of device historical data and prediction of device status changes refers to real-time collection of device data, caching the collected data, and real-time calculation of the device load rate L and temperature deviation T e ;

[0010] Initialize the sampling frequency to the default base value f base , through the equipment load rate L and temperature deviation T e Dynamically adjust the data sampling frequency, the calculation formula is:

[0011] f=f base ×(1+k1·L+k2·|T e ∣),

[0012] Among them, f is the sampling frequency after dynamic adjustment, L is the current load rate of the device, |T e ∣ is the absolute value of the current temperature deviation, k1 and k2 are the weight factors of load rate and ambient temperature;

[0013] By performing noise removal and outlier processing on the real-time device data obtained by dynamic sampling frequency and the collected device historical data;

[0014] The device data includes voltage, temperature, load, device alarm and fault code;

[0015] Build an LSTM network model and train it using historical device data. Use the gradient descent method for model training, select mean square error as the loss function for predicting future device state values, calculate the gradient of the loss function model parameters using the backpropagation algorithm, and use the Adam optimizer to update the model parameters. Stop iteration when the model loss no longer decreases significantly during continuous iterations. Output the model parameters to obtain a trained LSTM network model.

[0016] Input real-time device data into the trained LSTM network model to obtain the future device state value;

[0017] Calculate the equipment failure risk value based on the future equipment status value:

[0018]

[0019] Where R is the fault risk value, N is the length of the sampling time series, ΔT eis the temperature deviation of the device, S pred,i is the device state value predicted at the i-th sampling moment, L i is the equipment load rate at the i-th sampling moment, S max,i is the device safety threshold at the i-th sampling moment.

[0020] As a preferred solution of the method for actively reporting device status information described in the present invention, the optimization of the data uploaded in each period refers to calculating the similarity of the data transmitted in each reporting period by weighted fusion based on the adjusted data reporting period and fusing the data, and compressing the fused data using a Huffman coding algorithm;

[0021] Prioritize the compressed data according to importance: classify equipment failure information and predicted abnormal status data as high-priority data;

[0022] Classify equipment periodic status data into medium priority data;

[0023] Classify redundant status data and log information as low-priority data.

[0024] As a preferred solution of the method for actively reporting device status information described in the present invention, wherein: the real-time monitoring of network operation status refers to using key indicators of network operation to calculate the network congestion level and generate a comprehensive performance value of the transmission path, and calculating the congestion level of the path W i :

[0025]

[0026] Among them, W i is the congestion level of the i-th path, B i is the delay of the i-th path, D i is the bandwidth utilization of the i-th path, A i is the packet loss rate of the i-th path;

[0027] Calculate the comprehensive performance value S of the i-th path i , the formula is:

[0028] S i =w1·B i -w2·D i -w3·A i +w4·(1-W i ),

[0029] Among them, S i is the comprehensive performance value of the i-th path, w1, w2, w3, w4 are dynamically adjusted weight factors, and the path comprehensive performance value S iThe transmission paths are ranked in descending order to generate a candidate path priority list E;

[0030] According to the matching of data priority and path priority, high-priority data will be allocated to the optimal path P1, medium-priority data will be allocated to the sub-optimal path P2, and low-priority data will be allocated to the remaining backup path.

[0031] As a preferred scheme of the device state information active reporting method, wherein: the optimized data is uploaded to the server according to the network state, which means that the priority score R(E) of each path is calculated combined with the path priority list E, and the calculation formula is:

[0032]

[0033] Wherein, n is the total number of paths in the priority list, R ank (E) is the ranking of the path in the priority list;

[0034] The fitness value F of each path is calculated using the genetic algorithm, and the calculation formula is:

[0035]

[0036] Wherein, J i is the remaining energy of the path;

[0037] According to the mean and standard deviation of the fitness of the candidate path set, a dynamic threshold Q i is set, and the path subset P i is selected with a fitness value greater than the screening threshold Q GA ;

[0038] Each path P i is represented as a performance characteristic vector x i =[B i , D i , A i , J i ], the particle swarm size is set to N pop , the speed v i is a random value, and the particle fitness value F(x i ) is:

[0039]

[0040] The initial individual optimal solution of each particle is set to the particle's own position P best =x i , and the particle with the maximum fitness value is selected from the particle swarm as the initial global optimal solution g best :

[0041] g best= argmax{F(x i )},

[0042] The position of the global optimal solution g best is taken as the target node x goal ;

[0043] The path particle is driven to the target node x goal = g best by using the attractive potential field, and the performance characteristic function is obtained by integrating the performance value S i :

[0044]

[0045] wherein, is a nonlinear function;

[0046] The partial derivative of the formula H(S i ) is taken to obtain the gradient calculation formula:

[0047]

[0048] The partial derivatives of all components are arranged in vector form:

[0049]

[0050] The attractive potential field formula is defined as:

[0051]

[0052] wherein, F att (x i ) is the result value of the attractive potential field, x goal is the position of the target node, ||x goal -x i || represents the Euclidean distance between the current particle and the target point, is the gradient of the path performance characteristic function;

[0053] The repulsive potential field formula is defined as:

[0054]

[0055] wherein F rep (x i ) represents the repulsive force of the potential field, x c is the position of the congestion node, U0 is the action range of the repulsive potential field, and k rep is the repulsive force coefficient;

[0056] The attractive potential field and the repulsive potential field are combined to calculate the total moving direction of the path particle, and the calculation formula is:

[0057] Ftotal (x i )=F att (x i )+F rep (x i ),

[0058] Among them F total (x i ) is the total potential field value;

[0059] The potential field model is introduced to iteratively optimize the speed and position of the particle swarm to form the final optimization path. The calculation formula is:

[0060] v i (t+1)=ω·v i (t)+c1·r1·(Y best -x i (t))+c2·r2·(g best -x i (t))+F total (x i (t))

[0061] x i (t+1)=x i (t)+v i (t+1),

[0062] Among them, v i (t+1) is the velocity of particle i in the t+1th iteration, v i (t) is the velocity of particle i in the tth iteration, x i (t) is the position of particle i in the tth iteration, F total (x i (t) is the total potential field value of particle i in the tth iteration, ω is the inertia factor, c1, c2 are learning factors, r1, r2 are random numbers between [0,1], and Y best is the particle’s own historical optimal solution, g best is the global optimal solution, t is the number of iterations;

[0063] Each particle calculates a new fitness value F(x i ), and update:

[0064] The individual optimal solution is:

[0065] Y best =arg max{F(x i ),p best},

[0066] The global optimal solution is:

[0067] g best=arg max{F(x i )},

[0068] According to the calculated potential field value F total (x i ) and the particle swarm optimized path, select different transmission protocols, and establish a candidate list of protocols based on the data type:

[0069] High priority data selection F total (x i ) The largest path, choose TCP protocol for transmission;

[0070] Medium-priority data uses a suboptimal path and the UDP protocol for transmission;

[0071] Low-priority data selects the remaining available paths and is transmitted using the subnet broadcast protocol.

[0072] As a preferred solution of the method for actively reporting device status information described in the present invention, the storage device status information refers to hierarchical storage according to the type of uploaded data, compression before data storage, and generation of a unique index value for each data record. After storage is completed on the server side, the stored data is analyzed.

[0073] In a second aspect, the present invention provides a system for actively reporting device status information, including a data acquisition and prediction module for outputting high-quality device status data and prediction results;

[0074] Data processing and optimization module, used to compress and optimize device data and divide it into different priority categories;

[0075] Network status monitoring and path optimization module, used to generate candidate path lists and dynamically optimized transmission paths;

[0076] The data transmission module is used to select the appropriate protocol after the optimized path for data transmission;

[0077] The data storage and analysis module is used to store device status information and support fast retrieval and data analysis.

[0078] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the method for actively reporting device status information as described in the first aspect of the present invention is implemented.

[0079] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of a method for actively reporting device status information as described in the first aspect of the present invention.

[0080] The beneficial effects of the present invention are as follows: by dynamically sampling data, compression and priority division are performed, the data collection efficiency is optimized, the importance of data transmission is ensured, the data reporting cycle is dynamically adjusted according to the fault risk value, the amount of invalid data transmission is reduced, resource utilization is optimized, and the efficiency and pertinence of data reporting are improved. By real-time monitoring of the network operation status and optimizing path selection, uploading of duplicate data is avoided, the network transmission burden is reduced, and high-priority data is ensured to be allocated to the optimal path. The use of dynamic adjustment mechanism and intelligent algorithm improves the real-time, reliability and resource utilization of data upload. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0082] Figure 1 This is a flow chart of a method for actively reporting device status information in Example 1.

[0083] Figure 2 This is a structural diagram of a system for actively reporting device status information in Example 1. DETAILED DESCRIPTION

[0084] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0085] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0086] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0087] Example 1, with reference to Figure 1 and Figure 2 , which is the first embodiment of the present invention, provides a method for actively reporting device status information. The method for actively reporting device status information includes the following steps:

[0088] S1. Collect historical data of the equipment, predict the status of the equipment, and dynamically adjust the reporting cycle based on the prediction results;

[0089] Specifically, the device data is collected in real time, the collected data is cached, and the device load rate L and temperature deviation T are calculated in real time. e :

[0090]

[0091] T e =T current -T optimal ,

[0092] Among them, P current is the current power of the device, P max is the rated maximum power of the device, provided by the technical parameters of the device, T current Current device operating temperature, T optimal The optimal operating temperature for the equipment;

[0093] Initialize the sampling frequency to the default base value f base , through the equipment load rate L and temperature deviation T e Dynamically adjust the data sampling frequency, the calculation formula is:

[0094] f=f base ×(1+k1·L+k2·|T e ∣),

[0095] Where f is the dynamically adjusted sampling frequency, L is the current load rate of the device, ranging from [0,1], where L = 0 means the device is completely unloaded, and L = 1 means the device is fully loaded. e ∣ is the absolute value of the current temperature deviation, k1 and k2 are the weight factors of load rate and ambient temperature, which are obtained by fitting historical data;

[0096] By performing noise removal and outlier processing on the real-time device data obtained by dynamic sampling frequency and the collected device historical data;

[0097] The device data includes voltage, temperature, load, device alarm and fault code;

[0098] Build an LSTM network model and train it using historical device data. Use the gradient descent method for model training, select mean square error as the loss function for predicting future device state values, calculate the gradient of the loss function model parameters using the backpropagation algorithm, and use the Adam optimizer to update the model parameters. Stop iteration when the model loss no longer decreases significantly during continuous iterations. Output the model parameters to obtain a trained LSTM network model.

[0099] Input real-time device data into the trained LSTM network model to obtain the future device state value;

[0100] Based on the future equipment status value, the linear weighted and normalized failure risk assessment formula design method is used to calculate the basic equipment failure risk value R'. The calculation formula is:

[0101]

[0102] Among them, T trend The state change trend is obtained by performing differential calculation on the historical device state value and the predicted device state value. env is the environmental condition influencing factor, which is calculated through the original data of environmental conditions (such as temperature and humidity). θ1 and μ2 are the weight factors of state change trend and environmental conditions, respectively, which are obtained through the loss function of historical data. S pred To predict the device status value, S max is the device status threshold, obtained using statistical analysis based on historical device data;

[0103] The improved integrated nonlinear dynamic modeling and normalized calculation method is used to improve the equipment failure risk value R. The calculation formula is:

[0104]

[0105] Where R is the fault risk value, N is the length of the sampling time series, ΔT e is the temperature deviation of the device, S pred,i is the device state value predicted at the i-th sampling moment, L i is the equipment load rate at the i-th sampling moment, S max,i is the device safety threshold at the i-th sampling moment, which is obtained by statistical analysis based on the device's historical data.

[0106] Compared with the original fault risk value formula, the improved formula is more detailed and comprehensive in mathematical expression, and can more accurately express the complex interactive relationship between equipment operating status, environmental factors and fault risks. Especially in scenarios with complex data and long time series, the improved formula's nonlinear modeling ability and time series processing ability have obvious advantages.

[0107] The design of an improved formula is necessary, mainly because the original formula is difficult to reflect the complex dynamic relationship of the equipment status. The improved formula significantly enhances the modeling capability by introducing functions such as logarithms and square roots. The operating status of modern equipment often depends on the dynamic analysis of time series data. The improved formula effectively integrates historical data.

[0108] The strategy of dynamically adjusting the sampling frequency optimizes data collection efficiency, and the improved formula models the interactive relationship between equipment status, environmental conditions, and failure risk in more detail, providing a reliable basis for failure risk assessment.

[0109] Furthermore, dynamically adjusting the reporting cycle based on the prediction results refers to dynamically adjusting the reporting cycle based on the equipment failure risk value. The risk threshold Z is set by the standard deviation of the equipment's historical data. If the failure risk value R is greater than or equal to the set failure risk threshold Z, the equipment failure risk is considered high, and the reporting cycle is shortened, thereby increasing the frequency of reporting equipment status information.

[0110] If the fault risk value R is less than the set fault risk threshold Z, the device failure risk is considered low, the reporting cycle is extended, and the data transmission frequency is reduced.

[0111] The dynamic adjustment mechanism can adapt to changes in equipment operating status in real time. It can not only quickly capture key data when the risk is high, but also save transmission resources when the risk is low, avoiding the disadvantages of fixed reporting cycles, improving the accuracy and flexibility of equipment operating status monitoring, and providing more comprehensive support for fault detection and early warning.

[0112] S2. Optimize the data uploaded in each cycle and monitor the network operation status in real time;

[0113] Specifically, optimizing the data uploaded in each cycle means calculating the similarity of the data transmitted in each reporting cycle through weighted fusion based on the adjusted data reporting cycle, fusing the data, and compressing the fused data using the Huffman coding algorithm.

[0114] Prioritize the compressed data according to importance: classify equipment failure information and predicted abnormal status data as high-priority data;

[0115] Classify equipment periodic status data into medium priority data;

[0116] Classify redundant status data and log information as low-priority data.

[0117] The use of data fusion effectively reduces redundant data transmission, optimizes bandwidth utilization, and improves data utilization. This method has significant advantages in high-data-volume and complex network environments. Prioritization and dynamic adjustment strategies ensure the real-time performance of high-priority data, while reasonably scheduling medium and low-priority data, optimizing data transmission efficiency. The dynamic adjustment mechanism smoothes changes in data transmission load and avoids system instability caused by fixed upload strategies.

[0118] Furthermore, real-time monitoring of network operation status refers to using key indicators of network operation (key indicators are bandwidth utilization, packet loss rate, and delay) to calculate the network congestion level and generate a comprehensive performance value of the transmission path, and calculate the congestion level of the path W i :

[0119]

[0120] Among them, W i is the congestion level of the i-th path, B i is the delay of the i-th path, D i is the bandwidth utilization of the i-th path, A i is the packet loss rate of the i-th path;

[0121] Calculate the comprehensive performance value S of the i-th path i , the formula is:

[0122] S i =w1·B i -w2·D i -w3·A i +w4·(1-W i ),

[0123] Among them, S i is the comprehensive performance value of the i-th path, w1, w2, w3, and w4 are dynamically adjusted weight factors. The weight factors are dynamically calculated from historical data through a regression model. When network congestion increases, w3 (packet loss rate weight) is increased and w1 (bandwidth weight) is reduced. When the network is unobstructed, w1 and w4 are increased to give priority to efficient paths.

[0124] According to the comprehensive performance value S of the path i Sort the transmission paths in descending order to generate a candidate path priority list E:

[0125] E=[P1,P2,…,P n ],

[0126] Among them, P1 is the path with the highest performance value and is the optimal path, P2,…,P n The backup paths have decreasing comprehensive performance values.

[0127] Based on the matching of data priority and path priority, high-priority data will be allocated to the path with the best performance, P1; medium-priority data will be allocated to the suboptimal path, P2; and low-priority data will be allocated to the remaining backup paths.

[0128] The dynamic calculation of path comprehensive performance value, real-time adjustment of path selection logic according to network state, ensures that the optimal performance path can still be selected under network congestion, improves the reliability and efficiency of data transmission path, reduces the influence of network fluctuation on data transmission, the precise matching strategy of data priority and path priority significantly improves the transmission real-time performance and reliability of high-priority data, optimizes the overall transmission efficiency, continuously monitors the network state and dynamically adjusts the strategy, so that the system can quickly respond to network changes, and ensures the stability and reliability of data transmission.

[0129] S3, uploading the optimized data to the server according to the network state, and storing the device state information;

[0130] According to the network state, the optimized data is uploaded to the server, which means that the priority score R(E) of each path is calculated by combining the path priority list E, and the calculation formula is:

[0131]

[0132] Wherein, n is the total number of paths in the priority list, R ank (E) is the ranking of the path in the priority list, (starting from 1);

[0133] The fitness value F of each path is calculated using genetic algorithm, and the calculation formula is:

[0134]

[0135] Wherein, J i is the remaining energy of the path, and the historical energy consumption data (such as the number of sent data packets, the number of received data packets) is used to train the machine learning model to obtain the remaining energy;

[0136] According to the mean and standard deviation of the fitness of the candidate path set, set a dynamic threshold Q i , select the path subset P i with fitness value greater than the screening threshold Q GA ;

[0137] Each path P i is represented as a performance characteristic vector x i =[B i , D i , A i , J i ], set the particle swarm size as N pop , the speed v i is a random value, and the particle fitness value F(x i ) is:

[0138]

[0139] Set the initial individual optimal solution of each particle as the particle's own position P best = x i , Select the particle with the maximum fitness value from the particle swarm as the initial global optimal solution g best :

[0140] g best = argmax{F(x i )},

[0141] Take the position of the global optimal solution g best as the target node x goal ;

[0142] Use the attractive potential field to drive the path particles to move closer to the target node x goal = g best , and obtain the performance characteristic function by synthesizing the performance value S i :

[0143]

[0144] wherein, is a nonlinear function;

[0145] Take the partial derivative of the formula H(S i ) to obtain the gradient calculation formula:

[0146]

[0147] Arrange the partial derivatives of all components into vector form:

[0148]

[0149] Define the attractive potential field formula as:

[0150]

[0151] wherein, F att (x i ) is the result value of the attractive potential field, x goal is the position of the target node, represents the path target point that the particle needs to approach, ||x goal -x i || represents the Euclidean distance between the current particle and the target point, is the gradient of the path performance characteristic function, indicating the performance change trend of the path point, is a logarithmic adjustment function of the gradient size, sin(||x goal -x i || 3 ) represents the oscillation characteristics of the attractive force, is the normalized distance function, which is the smooth change of the distance between the particle and the target node;

[0152] The repulsive potential field formula is defined as:

[0153]

[0154] Among them F rep (x i ) represents the force of the repulsive potential field, that is, particle x i In the process of path optimization, the degree of rejection by congested nodes, x c is the location of the congested node, which is obtained from the path performance characteristics. U0 is the range of the repulsive potential field, and the repulsive force range is dynamically controlled by the performance difference between the path particles and the congested nodes. rep is the repulsion coefficient, which is set according to the degree of network congestion;

[0155] Combine the attractive potential field and the repulsive potential field to calculate the total moving direction of the path particles. The calculation formula is:

[0156] F total (x i )=F att (x i )+F rep (x i )

[0157] Among them F total (x i ) is the total potential field value, which determines the direction of particle movement;

[0158] The potential field model is introduced to iteratively optimize the speed and position of the particle swarm to form the final optimization path. The calculation formula is:

[0159] v i (t+1)=ω·v i (t)+c1·r1·(Y best -x i (t))+c2·r2·(g best -x i (t))+F total (x i (t))

[0160] x i (t+1)=x i (t)+v i (t+1),

[0161] Among them, v i (t+1) is the velocity of particle i in the t+1th iteration, v i (t) is the velocity of particle i in the tth iteration, x i(t) is the position of particle i in the tth iteration, F total (x i best (t) is the total potential field value of particle i in the tth iteration, ω is the inertia factor, c1, c2 are learning factors, r1, r2 are random numbers between [0, 1], Y best is the history optimal solution of the particle itself, g i is the global optimal solution, and t is the iteration number;

[0162] Each particle calculates a new fitness value F(x best ) and updates:

[0163] The individual optimal solution is:

[0164] Y i = arg max{F(x best ), p

[0165] The global optimal solution is:

[0166] g best = arg max{F(x i )},

[0167] According to the calculated potential field value F total (x i ) and the optimized path of the particle swarm, different transmission protocols are selected, and a protocol candidate list is established based on the data type:

[0168] High priority data selects the path with the maximum F total (x i ), and selects TCP protocol for transmission;

[0169] Medium priority data selects the suboptimal path, and selects UDP protocol for transmission;

[0170] Low priority data selects the remaining available path, and uses subnet broadcast protocol for transmission.

[0171] Dynamic path priority ranking combined with fitness calculation and potential field model comprehensively considers path performance and network congestion degree, realizes global optimal path selection, makes path selection more accurate, flexible and adaptive, can respond to network dynamic changes in real time, improves the reliability of data upload, dynamically selects protocols according to data types, significantly improves the transmission efficiency, and is suitable for complex network environment, especially in the multi-path, high data volume and dynamic network state, which shows significant advantages.

[0172] ​Furthermore, the storage device status information refers to hierarchical storage based on the type of uploaded data, compression before data storage, and generation of a unique index value for each data record. After storage is completed on the server side, the stored data is analyzed.

[0173] Tiered storage allows high-priority data to be stored and accessed first, improving the management efficiency and access speed of important data, reducing the waste of storage resources, and improving the storage management efficiency of the system. Data analysis enables the full utilization of stored data, and an efficient data analysis framework enhances the intelligence level of the system.

[0174] This embodiment also provides a system for actively reporting device status information, including:

[0175] Data acquisition and prediction module, used to output high-quality equipment status data and prediction results;

[0176] Data processing and optimization module, used to compress and optimize device data and divide it into different priority categories;

[0177] Network status monitoring and path optimization module, used to generate candidate path lists and dynamically optimized transmission paths;

[0178] The data transmission module is used to select the appropriate protocol after the optimized path for data transmission;

[0179] The data storage and analysis module is used to store device status information and support fast retrieval and data analysis.

[0180] This embodiment also provides a computer device, which is suitable for a method for actively reporting device status information, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement a method for actively reporting device status information as proposed in the above embodiment.

[0181] The computer device may be a terminal, comprising a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a button, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse.

[0182] This embodiment also provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for actively reporting device status information as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0183] In summary, the present invention optimizes data acquisition efficiency by dynamically sampling data, performing compression and priority division, ensuring the importance of data transmission, dynamically adjusting the data reporting cycle according to the fault risk value, reducing the amount of invalid data transmission, optimizing resource utilization, and improving the efficiency and pertinence of data reporting. By real-time monitoring of the network operation status and optimizing path selection, uploading of duplicate data is avoided, the network transmission burden is reduced, and high-priority data is ensured to be allocated to the optimal path. The use of dynamic adjustment mechanisms and intelligent algorithms improves the real-time, reliability and resource utilization of data upload.

[0184] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for actively reporting device status information, characterized in that: include: Collect historical data of equipment, predict the status of equipment, and dynamically adjust the reporting cycle based on the prediction results; Optimize the data uploaded in each cycle and monitor the network operation status in real time; Upload optimized data to the server according to the network status and store device status information; The collecting of historical data of the equipment and the prediction of the state change of the equipment refer to collecting equipment data in real time, caching the collected data, and calculating the equipment load rate L and temperature deviation T in real time. e ; Initialize the sampling frequency to the default base value f base , through the equipment load rate L and temperature deviation T e Dynamically adjust the data sampling frequency, the calculation formula is: f=f base ×(1+k1·L+k2·∣T e ∣), Among them, f is the sampling frequency after dynamic adjustment, L is the current load rate of the device, |T e ∣ is the absolute value of the current temperature deviation, k1 and k2 are the weight factors of load rate and ambient temperature; By performing noise removal and outlier processing on the real-time device data obtained by dynamic sampling frequency and the collected device historical data; The device data includes voltage, temperature, load, device alarm and fault code; Build an LSTM network model and train it using historical device data. Use the gradient descent method for model training, select mean square error as the loss function for predicting future device state values, calculate the gradient of the loss function model parameters using the backpropagation algorithm, and use the Adam optimizer to update the model parameters. Stop iteration when the model loss no longer decreases significantly during continuous iterations. Output the model parameters to obtain a trained LSTM network model. Input real-time device data into the trained LSTM network model to obtain the future device state value; Calculate the equipment failure risk value based on the future equipment status value: Where R is the fault risk value, N is the length of the sampling time series, ΔT e is the temperature deviation of the device, S pred,i is the device state value predicted at the i-th sampling moment, L i is the equipment load rate at the i-th sampling moment, S max,i is the device safety threshold at the i-th sampling moment; Dynamically adjusting the reporting cycle based on the prediction results refers to dynamically adjusting the reporting cycle based on the equipment failure risk value. The risk threshold Z is set by the standard deviation of the equipment's historical data. If the failure risk value R is greater than or equal to the set failure risk threshold Z, the equipment failure risk is considered high, and the reporting cycle is shortened, thereby increasing the frequency of reporting equipment status information. If the fault risk value R is less than the set fault risk threshold Z, the device failure risk is considered low, the reporting cycle is extended, and the data transmission frequency is reduced.

2. A method for actively reporting device status information according to claim 1, characterized in that: Optimizing the data uploaded in each period refers to calculating the similarity of the data transmitted in each reporting period through weighted fusion based on the adjusted data reporting period and fusing the data, and compressing the fused data using the Huffman coding algorithm; Prioritize the compressed data according to importance: Classify equipment failure information and predicted abnormal status data as high-priority data; Classify equipment periodic status data into medium priority data; Classify redundant status data and log information as low-priority data.

3. The method for actively reporting device status information according to claim 2, characterized in that: The real-time monitoring of network operation status refers to using key indicators of network operation to calculate the network congestion level and generate a comprehensive performance value of the transmission path, and calculating the congestion level of the path W i : Among them, W i is the congestion level of the i-th path, B i is the delay of the i-th path, D i is the bandwidth utilization of the i-th path, A i is the packet loss rate of the i-th path; Calculate the comprehensive performance value S of the i-th path i , the formula is: S i =w1·B i -w2·D i -w3·A i +w4·(1-W i ), Among them, S i is the comprehensive performance value of the i-th path, w1, w2, w3, w4 are dynamically adjusted weight factors, and the path comprehensive performance value S i Sort the transmission paths in descending order to generate a candidate path priority list E; Based on the matching of data priority and path priority, high-priority data will be allocated to the path with the best performance, P1; medium-priority data will be allocated to the suboptimal path, P2; and low-priority data will be allocated to the remaining backup paths.

4. A method for actively reporting device status information according to claim 3, characterized in that: Uploading the optimized data to the server according to the network status refers to calculating the priority score R(E) of each path in combination with the path priority list E. The calculation formula is: Where n is the total number of paths in the priority list, R ank (E) is the ranking of the path in the priority list; Use the genetic algorithm to calculate the fitness value F for each path. The calculation formula is: Among them, J i is the residual energy of the path; According to the fitness mean and standard deviation of the candidate path set, a dynamic threshold Q is set i , select the fitness value greater than the screening threshold Q i Form a path subset P GA ; For each path P i Expressed as performance characteristic vector x i =[B i , D i , A i , J i ], set the particle swarm size to N pop , speed v i is a random value, the particle fitness value F(x i )for: Set the initial individual optimal solution of each particle to the particle's own position P best =x i , select the particle with the largest fitness value from the particle swarm as the initial global optimal solution g best : g best =argmax{F(x i )}, The global optimal solution g best The location of the target node x goal ; Use the attractive potential field to drive the path particles toward the target node x goal =g best Closer, through the comprehensive performance value S i Get the performance characteristic function: in, It is a nonlinear function; For the formula H(S i ) to find the partial derivative and obtain the gradient calculation formula: Arrange the partial derivatives of all components into vector form: The attractive potential field formula is defined as: Among them, F att (x i ) The resulting value of the attractive potential field, x goal is the position of the target node, ||x goal -x i || represents the Euclidean distance between the current particle and the target point, is the gradient of the path performance characteristic function; The repulsive potential field formula is defined as: Among them F rep (xi) represents the force of the repulsive potential field, x c is the location of the congested node, U0 is the range of the repulsive potential field, k rep is the repulsion coefficient; Combine the attractive potential field and the repulsive potential field to calculate the total moving direction of the path particles. The calculation formula is: F total (x i )=F att (x i )+F rep (x i ), Among them F total (x i ) is the total potential field value; The potential field model is introduced to iteratively optimize the speed and position of the particle swarm to form the final optimization path. The calculation formula is: v i (t+1)=ω·v i (t)+c1·r1·(Y best -x i (t))+c2·r2·(g best -x i (t))+F total (x i (t)) x i (t+1)=x i (t)+v i (t+1), Among them, v i (t+1) is the velocity of particle i in the t+1th iteration, v i (t) is the velocity of particle i in the tth iteration, x i (t) is the position of particle i in the tth iteration, F total (x i (t)) is the total potential field value of particle i in the tth iteration, ω is the inertia factor, c1, c2 are learning factors, r1, r2 are random numbers between [0,1], and Y best is the particle’s own historical optimal solution, g best is the global optimal solution, t is the number of iterations; Each particle calculates a new fitness value F(x i ), and update: The individual optimal solution is: Y best =arg max{F(x i ),p best }, The global optimal solution is: g best =arg max{F(x i )}, According to the calculated potential field value F total (x i ) and the particle swarm optimized path, select different transmission protocols, and establish a candidate list of protocols based on the data type: High priority data selection F total (x i ) The largest path, choose TCP protocol for transmission; Medium-priority data uses a suboptimal path and the UDP protocol for transmission; Low-priority data selects the remaining available paths and is transmitted using the subnet broadcast protocol.

5. A method for actively reporting device status information according to claim 4, characterized in that: The storage device status information refers to hierarchical storage based on the type of uploaded data, compression before data storage, and generation of a unique index value for each data record. After storage is completed on the server side, the stored data is analyzed.

6. A system for actively reporting device status information, based on a method for actively reporting device status information according to any one of claims 1 to 5, characterized in that: include, Data acquisition and prediction module, used to output high-quality equipment status data and prediction results; Data processing and optimization module, used to compress and optimize device data and divide it into different priority categories; Network status monitoring and path optimization module, used to generate candidate path lists and dynamically optimized transmission paths; The data transmission module is used to select the appropriate protocol after the optimized path for data transmission; The data storage and analysis module is used to store device status information and support fast retrieval and data analysis.

7. A computer device comprising: memory and processor; The memory stores a computer program, characterized in that when the processor executes the computer program, the steps of the method for actively reporting device status information according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a method for actively reporting device status information according to any one of claims 1 to 5 are implemented.

Images