Intelligent equipment remote cooperative control method and system based on Internet of Things

Abstract

The invention relates to the technical field of remote control, in particular to an intelligent device remote cooperative control method and system based on the Internet of Things, and the method comprises the following steps: monitoring and adjusting the CPU utilization rate and memory usage of an intelligent device based on an Internet of Things platform, optimizing the data collection and network resource distribution, and improving the intelligent device remote cooperative control efficiency. By analyzing equipment performance and a network load balance index, equipment connection priority ranking and task load management are realized, and task execution and connection frequency is dynamically adjusted, so that real-time task requirements are met, and system efficiency is improved. According to the method, the performance index of the equipment is monitored in real time, the data transmission interval is dynamically adjusted, the data processing speed is remarkably increased, resource waste is reduced, efficient use of network resources is ensured through real-time load analysis and optimization of network nodes, the problem of data delay caused by network congestion is avoided, and in the aspect of equipment connection management, the data transmission efficiency is improved. An intelligent priority decision mechanism allows the key device to quickly access the network in an emergency.

Description

Technical Field

[0001] The present invention relates to the field of remote control technology, and in particular to a remote collaborative control method and system for intelligent devices based on the Internet of Things. Background Art

[0002] Remote control is a technology that uses communication technology to monitor and operate remote devices or systems. With the rapid development of the Internet of Things, wireless communication technology and the Internet, remote control technology has been widely used in smart homes, industrial automation, environmental monitoring, health management and other fields. Through remote control, users can monitor and adjust devices in real time through the network or other communication means at a distance from the target device, which improves the convenience and safety of operation.

[0003] Among them, the remote collaborative control method of smart devices in the Internet of Things is an application of remote control technology, which aims to enable different smart devices to achieve remote collaborative work through the Internet of Things technology. This method allows multiple smart devices to be interconnected, share information and interact in real time through the network to achieve unified management and control of smart devices. Its main purpose is to enhance the collaboration and automation between devices, improve the operating efficiency and reliability of smart devices, and is widely used in scenarios such as smart home systems, equipment management of industrial production lines, and environmental monitoring.

[0004] The existing technology has a fixed data update frequency and cannot be quickly adjusted to respond to emergencies, resulting in an inability to respond immediately when the device status changes rapidly, affecting the accuracy of decision-making. In terms of network management, the existing technology fails to optimize the network load in real time and cannot effectively respond to load fluctuations, which is prone to data transmission delays or losses. In addition, the priority management of device access lacks a dynamic adjustment mechanism, and cannot ensure that key tasks receive priority processing when network resources are tight. This is particularly unfavorable in high-demand real-time monitoring and control scenarios, and can easily lead to delays or failures in important data processing, affecting the overall control stability and security. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a remote collaborative control method and system for intelligent devices based on the Internet of Things.

[0006] In order to achieve the above object, the present invention adopts the following technical scheme: a remote collaborative control method of intelligent devices based on the Internet of Things, comprising the following steps:

[0007] S1: Based on the IoT platform, the CPU utilization and memory usage of smart devices are monitored, and the status data of smart devices are regularly collected using the IoT communication interface. The data transmission interval is dynamically adjusted through real-time analysis of the performance of smart devices to obtain the device performance monitoring data set;

[0008] S2: Based on the device performance monitoring data set, obtain the bandwidth utilization rate, signal strength, and network response time of the intelligent device in the edge network, conduct data comparison on network parameters, identify and adjust the network nodes with the load reaching the upper limit, and obtain the network load balance index;

[0009] S3: Based on the network load balance index, analyze the CPU utilization rate and memory usage of the intelligent device, automatically prioritize connection requests according to the response time and signal strength of the network nodes connected by each intelligent device, mark the high-priority devices as having the priority connection right, and delay the connection of the remaining devices to obtain the access priority index;

[0010] S4: Based on the access priority index, evaluate the task queue length and memory usage rate of the connected devices, and adjust the task allocation ratio according to the data processing performance of the devices to obtain the task load management information.

[0011] The improvement of the present invention is that the device performance monitoring data set includes performance trends, resource occupancy rates, and status feedback information, the network load balance index includes load capacity and adjustment response results, the access priority index includes weight levels, sorting logics, and response priorities, and the task load management information includes processing capabilities, queue optimization results, and task response speeds.

[0012] The improvement of the present invention is that the acquisition steps of the device performance monitoring data set are specifically as follows:

[0013] S111: Based on the Internet of Things platform, use the communication interface to collect the status data of the intelligent device, monitor the CPU utilization rate, memory usage rate, and current network bandwidth occupancy, identify the missing data and fill it to obtain the resource utilization time series;

[0014] S112: According to the resource utilization time series, analyze the status data of the intelligent device in consecutive operation cycles, calculate its change rate within the time, compare it with the known stable change interval, and mark the time periods exceeding the fluctuation benchmark to obtain the stability fluctuation trend;

[0015] S113: For the stability fluctuation trend, analyze the fluctuation degree interval in the current cycle and the fluctuation change level in the historical stable cycle, and use the formula:

[0016]

[0017] Obtain the performance change intensity DE of the i-th device i , and dynamically adjust the data transmission interval to obtain the device performance monitoring data set, where Δce i is the CPU utilization rate change value of the i-th device, and Δme iis the change value of the memory occupancy rate of the i-th device, be i is the bandwidth fluctuation frequency of the i-th device, re i is the response time lag factor of the i-th device, de i is the data interaction density factor of the i-th device, qe ik is the concurrent device impact coefficient of the i-th device in the k-th time period, n de is the total number of time periods.

[0018] The improvement of the present invention is that the obtaining step of the network load balancing index is specifically as follows:

[0019] S211: According to the device performance monitoring data set, obtain the bandwidth utilization rate, signal strength and network response time corresponding to the intelligent device in the edge network, align the time series of network parameters, and group according to the network node positions to establish the initial network load data;

[0020] S212: Call the initial network load data, compare the load of each network node with the average load of all nodes, and use the formula:

[0021]

[0022] Calculate the network resource transfer performance RT j , identify and adjust the network nodes with the load reaching the upper limit to obtain the network load balancing index, where BT j represents the bandwidth utilization rate of the current node j, BT z represents the bandwidth utilization rate of other nodes z compared with the current node j, ST z represents the signal strength of other nodes z, TR z represents the network response time of other nodes z, α and β are the weights for adjusting the bandwidth difference and the ratio of signal strength to response time, n rt is the total number of nodes participating in the calculation.

[0023] The improvement of the present invention is that the obtaining step of the access priority index is specifically as follows:

[0024] S311: Based on the network load balancing index, analyze the CPU utilization rate and memory usage of the intelligent device, collect the network node data of each intelligent device, sort the connection requests of different devices to obtain the connection request sorting table;

[0025] S312: Call the connection request sorting table, and according to the response time and signal strength of the network node to which the device is connected, use the formula:

[0026]

[0027] Calculate the access priority score PR, and automatically sort the connection requests according to the score to obtain the device connection sorting result, where GR ij represents the response time of device i under node j, GR max represents the maximum response time of all devices under all nodes, GS ij represents the signal strength of device i under node j, GS max represents the maximum signal strength of all devices under all nodes, GL ij represents the network load balance index of device i under node j, GL max represents the maximum network load balance index of all devices under all nodes, N pr is the weighted average total number;

[0028] S313: According to the device connection sorting result, mark the high-priority devices as having the priority connection right, delay the connection of the remaining devices, and at the same time call the connection request sorting table to sort the delayed devices again to obtain the access priority index.

[0029] The improvement of the present invention is that the step of obtaining the task load management information is specifically as follows:

[0030] S411: Based on the access priority index, evaluate the real-time task queue data and the current memory usage data of the connected devices, compare the number of task requests with the set device request capacity, and compare the memory usage with the memory capacity to judge the current load level of the devices and obtain the device load judgment result;

[0031] S412: Based on the device load judgment result, use the formula:

[0032]

[0033] Calculate the task allocation offset value ΔGD, where CU represents the data processing performance parameter of the device, QD represents the current task request number of the device, RD represents the task response rate, MD represents the free memory capacity of the device, and K GD is the adjustment coefficient;

[0034] S413: According to the task allocation offset value, proportionally reallocate the device task request numbers, and combine the original task queue structure to reconstruct the scheduling information field to obtain the task load management information.

[0035] The improvement of the present invention also includes:

[0036] S5: Call the task load management information, analyze the task execution density and memory usage status according to the current working state of the intelligent device, dynamically adjust the connection time interval and task period to match the real-time task requirements, and synchronously adjust the device connection frequency to obtain the device connection control feedback result;

[0037] The device connection control feedback result includes connection adjustment data, periodic update information, and frequency synchronization metrics.

[0038] The improvement of the present invention is that the steps for obtaining the device connection control feedback result are specifically as follows:

[0039] Call the task load management information, based on the current working state of the intelligent device, compare the corresponding relationship between the task execution frequency and the memory usage rate, determine whether the task load exceeds the standard, and combine the number of intelligent device threads and the task cycle interval to determine the task load change trend;

[0040] Based on the task load change trend, analyze the matching degree between the task load and the connection period, and according to the changes in the task frequency and the device connection success rate, synchronously adjust the device connection frequency and the task cycle configuration to obtain the device connection control feedback result.

[0041] An intelligent device remote collaboration control system based on the Internet of Things, the system includes:

[0042] The device status monitoring module, based on the Internet of Things platform, monitors the CPU utilization rate and memory usage rate of the intelligent device, regularly collects the status data of the intelligent device, and dynamically adjusts the data transmission interval by real-time analyzing the performance of the intelligent device to obtain the device performance monitoring data set;

[0043] The network load management module, according to the device performance monitoring data set, obtains the bandwidth utilization rate, signal strength, and network response time of the intelligent device in the edge network, conducts data comparison on network parameters, identifies and adjusts the network nodes with the load reaching the upper limit to obtain the network load balance metrics;

[0044] The device access priority module, based on the network load balance metrics, analyzes the CPU utilization rate and memory usage of the intelligent device, and according to the response time and signal strength of the network node, automatically ranks the connection requests in priority, marks the high-priority devices as having the priority connection right, and delays the connection of the remaining devices to obtain the access priority metrics;

[0045] The intelligent task scheduling module, based on the access priority metrics, evaluates the task queue length and memory usage rate of the connected devices, and adjusts the task allocation ratio according to the data processing performance of the devices to obtain the task load management information;

[0046] The connection dynamic regulation module calls the task load management information, analyzes the task execution density and memory usage status, dynamically adjusts the connection time interval and the task cycle, matches the real-time task requirements, and synchronously adjusts the device connection frequency to obtain the device connection control feedback result.

[0047] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0048] In the present invention, by real-time monitoring the performance indicators of the device and dynamically adjusting the data transmission interval, the data processing speed is significantly improved and resource waste is reduced. Through real-time load analysis and optimization of network nodes, efficient use of network resources is ensured, and data delay problems caused by network congestion are avoided. In terms of device connection management, the intelligent priority determination mechanism allows critical devices to quickly access the network in emergency situations, enhancing the overall response ability of intelligent devices. Through intelligent analysis and adjustment of task allocation, the work coordination between devices is optimized, and the control efficiency and device usage stability are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 is the main step flow chart proposed by the present invention;

[0050] Figure 2 is the flow chart for obtaining the device performance monitoring data set in the present invention;

[0051] Figure 3 is the flow chart for obtaining the network load balancing index in the present invention;

[0052] Figure 4 is the flow chart for obtaining the access priority index in the present invention;

[0053] Figure 5 is the flow chart for obtaining the task load management information in the present invention;

[0054] Figure 6 is the flow chart for obtaining the device connection control feedback result in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0055] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0056] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. In addition, in the description of the present invention, "a plurality of" means two or more unless otherwise specifically defined.

[0057] Embodiment

[0058] Please refer to Figure 1 , the present invention provides a technical solution: an intelligent device remote collaborative control method based on the Internet of Things, comprising the following steps:

[0059] S1: Based on the Internet of Things platform, monitor the CPU utilization rate and memory usage of intelligent devices, regularly collect the status data of intelligent devices by using the Internet of Things communication interface, and dynamically adjust the data transmission interval through real-time analysis of the performance of intelligent devices to optimize the data collection efficiency, so as to obtain a device performance monitoring data set;

[0060] S2: According to the device performance monitoring data set, obtain the bandwidth utilization rate, signal strength and network response time of intelligent devices in the edge network, compare the network parameters, identify and adjust the network nodes with the load reaching the upper limit, optimize the network resource allocation, and obtain a network load balance index;

[0061] S3: Based on the network load balance index, analyze the CPU utilization rate and memory usage of intelligent devices, automatically prioritize the connection requests according to the response time and signal strength of the network nodes connected by each intelligent device, mark the high-priority devices as having the priority connection right, and delay the connection of the remaining devices to obtain an access priority index;

[0062] S4: Based on the access priority index, evaluate the task queue length and memory usage of the connected devices, adjust the task allocation ratio according to the data processing performance of the devices, optimize the device working load, and obtain task load management information;

[0063] S5: Invoke the task load management information, analyze the task execution density and memory usage according to the current working state of the intelligent device, dynamically adjust the connection time interval and task period, match the real-time task requirements, and synchronously adjust the device connection frequency to obtain a device connection control feedback result.

[0064] The device performance monitoring data set includes performance trends, resource occupancy rates, status feedback information, the network load balance index includes load capacity, adjustment response results, the access priority index includes weight levels, sorting logics, response priorities, the task load management information includes processing capabilities, queue optimization results, task response speeds, and the device connection control feedback result includes connection adjustment data, cycle update information, frequency synchronization indicators.

[0065] Please refer to Figure 2 , the specific steps for obtaining the device performance monitoring data set are as follows:

[0066] S111: Based on the Internet of Things platform, use the communication interface to collect the status data of intelligent devices, monitor the CPU utilization rate, memory usage rate, and current network bandwidth occupancy, identify the missing data and fill it to obtain the resource utilization time series;

[0067] Based on the Internet of Things platform, connect to the communication interface of the intelligent device, access it in the form of a serial port or Ethernet interface and identify the device number. After determining the hardware structure of the device, sequentially call the CPU usage rate, memory usage rate, and network bandwidth interface to collect status data. In each preset sampling period, record the above index data in seconds. Among them, the CPU utilization data is captured through the resource occupancy management function of the device's own operating system. For example, the Windows system calls the PerformanceMonitor data interface to obtain the process-level CPU usage percentage. If the initial value is 35% and the end value is 65%, then the CPU utilization change of this device within the interval can be obtained as 30%; the memory usage rate is obtained by the ratio of the memory page file call volume to the total available memory. If the start is 40% and the end is 75%, then the memory usage rate change is 35%; the network bandwidth is calculated by reading the network card traffic statistics value to calculate the average bandwidth occupancy. The collection start is 20Mbps, the end is 50Mbps, and the change amount is 30Mbps. At the same time, to identify data missing situations, the system judges whether the continuous time difference in the sampling timestamp sequence exceeds the set period ±5%. If 2 groups of data are missing in a sampling period, placeholders are inserted in the timeline and linear interpolation or moving average is started for filling. If the CPU at a certain point is 60% and 70% for the two adjacent points, the missing point will be estimated as 65%. The above sampling and repaired data form a resource utilization time series on the timeline. This series is used for subsequent fluctuation trend analysis and dynamic control decision-making. The resource utilization time series is finally stored in the database in the form of a one-dimensional vector. For example, if device 1 samples once every 30 seconds within 5 minutes, there will be 10 groups of records. The data storage structure is [timestamp, device number, CPU utilization rate, memory usage rate, bandwidth occupancy]. If the sampling time coincides or the offset exceeds 1 second during the continuous operation period, the corresponding sampling group is marked as an abnormal point to obtain a stable and continuous time series.

[0068] S112: According to the resource utilization time series, analyze the status data of the intelligent device during the continuous operation period, calculate its change rate within the time, compare it with the known stable change interval, and mark the time period exceeding the fluctuation benchmark to obtain the stability fluctuation trend;

[0069] According to the constructed resource utilization time series, the device operation data is divided into multiple consecutive periods according to the set sampling period. For example, 5 minutes is divided into 10 30-second periods. For each period, the change rates of three parameters, namely CPU utilization, memory usage, and network bandwidth, are calculated. The change rate is obtained by subtracting the value of the previous period from the current value and then dividing by the value of the previous period. For example, if the CPU utilization of a device rises from 65% to 75%, the change rate is (75% - 65%) / 65% = 15.38%. Similarly, if the memory usage drops from 75% to 70%, the change rate is (70% - 75%) / 75% = -6.67%. The three parameters are combined into a three-dimensional change vector, and then the Euclidean norm of the three-dimensional change vector generated for each period is calculated as the overall resource fluctuation intensity of this period. This value is compared with the preset stable fluctuation threshold range of the system. For example, when the set stable fluctuation range is [-10%, +10%], 15.38% will be recognized as exceeding the fluctuation benchmark range, and the system marks this period as an unstable time period. Then, the status of each period is marked to form a "stable / unstable" sequence for long-term trend modeling and statistical analysis. This stability fluctuation trend is further recorded as a scalar sequence and synchronized with the device operation log for update.

[0070] S113: For the stability fluctuation trend, analyze the fluctuation degree range in the current period and the fluctuation change level in the historical stable periods, and use the formula:

[0071]

[0072] Obtain the performance change intensity DE of the i-th device i , and dynamically adjust the data transmission interval to obtain the device performance monitoring data set, where Δce i is the CPU utilization change value of the i-th device, representing the change in CPU utilization of the device during the evaluation period, and Δme i is the memory occupancy change value of the i-th device, representing the change in memory occupancy of the device during the evaluation period, and be i is the bandwidth fluctuation frequency of the i-th device, used to reflect the fluctuation of network bandwidth usage of the device during the period, and re i is the response time lag factor of the i-th device, considering the impact of the delay of the device response time on performance evaluation, and de i is the data interaction density factor of the i-th device, reflecting the data interaction frequency and density between the device and other devices or systems in the network, and qe ik is the concurrent device impact coefficient of the i-th device in the k-th time period, measuring the impact of other devices on the performance of this device during the same time period, and n de is the total number of time periods, which is the total number of time segments for all devices;

[0073] For the obtained stability fluctuation trend data, select the status data of the current i-th device within a certain operation cycle. In this cycle, first obtain the change value Δce of its CPU utilization rate i , which is obtained by subtracting the maximum and minimum CPU values in the cycle and normalizing. For example, if the maximum CPU value in the cycle is 80% and the minimum value is 60%, the normalization is: (80 - 60) / 100 = 0.20, obtaining Δce i = 0.20; Similarly, for the change value Δme of the memory usage rate i , if the maximum value is 90% and the minimum value is 70%, then the change value after normalization is (90 - 70) / 100 = 0.20, obtaining Δme i = 0.20; The bandwidth fluctuation frequency be i is determined by judging the number of turning points of the bandwidth value curve per unit time. Suppose there are 6 bandwidth curve inflection points in this cycle, and the upper limit of the maximum fluctuation frequency supported by the device is 10 times. After normalization, it is 6 / 10 = 0.60, obtaining be i = 0.60; The response time lag factor re i is calculated based on the proportion of all sampled response values higher than the system standard value (such as the standard delay of 200 ms). For example, among 10 samplings, 3 are higher than 250 ms, and the lag threshold is set to 250 ms. After normalization, it is 3 / 10 = 0.30, obtaining re i = 0.30; The data interaction density factor de i is set as the proportion of the total number of data transmissions and receptions per unit time to the maximum expected interaction frequency. If the total number of interactions of the current device in the cycle is 200 times and the maximum designed interaction frequency is 400 times, after normalization, it is 200 / 400 = 0.50, obtaining de i = 0.50; The concurrent device impact coefficient qe ik represents the occupancy degree of other devices on the performance of the current device. Suppose the observed impact ratios in three sub-time periods are 0.40, 0.30, and 0.20 respectively. Summing them up gives:

[0074]

[0075] Substitute the above values into the formula:

[0076]

[0077] The calculation result shows that the performance change intensity of the i-th device in the current cycle is 0.2947. Comparing it with the preset performance fluctuation benchmark value (for example, the maximum allowable intensity under stable operation is 0.25), it is determined that it is in a state of excessive fluctuation. Therefore, it is necessary to adjust the sampling frequency and data transmission interval of the device. For example, extend the transmission cycle from the original 10 seconds to 20 seconds to reduce system interference, and synchronously record this change intensity and the adjustment strategy into the device performance monitoring dataset.

[0078] Please refer to Figure 3 , and the specific steps for obtaining the network load balance index are as follows:

[0079] S211: According to the device performance monitoring dataset, obtain the bandwidth utilization rate, signal strength, and network response time corresponding to the intelligent device in the edge network, align the time series of network parameters, and group them according to the network node location to establish the initial network load data;

[0080] Obtain the device performance monitoring dataset, parse each record in the dataset to extract the bandwidth utilization rate, signal strength, and network response time parameters corresponding to the intelligent device, align the time series of the three parameters based on the timestamp field. During the alignment process, group the records of different devices into the same time period in milliseconds, set a unified time benchmark of one sampling interval every 5 seconds, select multiple intelligent camera devices in the same office area in the actual scenario as the monitoring objects, synchronously extract the network parameters corresponding to the devices according to the time interval. For example, at time t1, the bandwidth of device A is 65%, the signal strength is -50dBm, and the response time is 20ms. At the same time, the values of device B are 60%, -48dBm, and 22ms respectively. Then, they are grouped into the same set of data in this time period. After completing the time registration of the network parameters of multiple devices, classify the nodes according to the installation location coordinates of the devices in the actual physical scenario. For example, define all devices in the southeast corner of a certain office area as node 1, perform mean processing on the three network parameters corresponding to multiple devices in the node, and respectively obtain the average bandwidth utilization rate, average signal strength, and average response time of the node in this time period, and use these as the network initial load data values of the node. Assume that a total of 6 devices are collected in node 1, then the average value of the bandwidth utilization rate can be expressed as (65 + 63 + 64 + 66 + 62 + 64) / 6 = 64%, and the other parameters are calculated in the same way for the mean value, and an initial parameter data table is constructed for subsequent calls. The node data will be used as the basic benchmark for subsequent load identification.

[0081] S212: Call the initial network load data, compare the load of each network node with the average load of all nodes, and use the formula:

[0082]

[0083] Calculate the network resource transfer performance RT j , identify and adjust the network nodes whose load reaches the upper limit to obtain the network load balance index, where BT j represents the bandwidth utilization rate of the current node j, which is an indicator to measure the current bandwidth usage of this node, BT z represents the bandwidth utilization rate of other nodes z compared with the current node j, which is used to determine the bandwidth usage difference between nodes, ST z represents the signal strength of other nodes z, which is an indicator to measure the communication efficiency and quality of the node, TR z represents the network response time of other nodes z, which is used to evaluate the speed of the node to process requests. The shorter the response time, the higher the efficiency of the node. α and β are the weights to adjust the ratio of the bandwidth difference and the signal strength to the response time, n rt is the total number of nodes participating in the calculation;

[0084] Assume the total number of nodes is n rt = 3. Take node j as the node to be calculated, and the comparison nodes are z1 and z2. The original values and normalized values of the network parameters of the nodes are as follows: the bandwidth utilization rate of node j is 65% (normalized value 1.0), the bandwidth utilization rate of node z1 is 60% (normalized value 0.5), and the bandwidth utilization rate of node z2 is 55% (normalized value 0.0); the signal strength of node z1 is -48dBm (normalized value 0.8), and the signal strength of z2 is -52dBm (normalized value 0.0); the response time of node z1 is 22ms (normalized value 0.4286), and the response time of z2 is 25ms (normalized value 0.0). Use the weight coefficients α = 0.6 and β = 0.4 and substitute them into the formula:

[0085]

[0086] Calculate the first group to get:

[0087] The second group is: (processed as 0), calculate:

[0088]

[0089] Get RT j = 0.8234. This result indicates that node j has a medium-level transfer ability in the comparison of resource parameters with its adjacent nodes. When the system sets the resource transfer benchmark value to 0.8, this node meets the scheduling adjustment condition, indicating that it has partial network load balancing ability. The higher the value, the more suitable the node is as the carrier or transfer target of network resources. Further, by horizontally sorting and comparing the RT values of all nodes, the nodes with RT values higher than the benchmark value can be screened as load scheduling nodes, thereby generating the network load balance index.

[0090] Please refer to Figure 4 , and the specific steps for obtaining the access priority metric are as follows:

[0091] S311: Based on the network load balancing metric, analyze the CPU utilization rate and memory usage of intelligent devices, collect the network node data of each intelligent device, sort the connection requests of different devices, and obtain a connection request sorting table;

[0092] By reading the real-time resource usage rate recorded in the device operating system or device management interface, collect the status data of the device under multiple network nodes. Among them, the response time is measured by the round-trip time of the network request, the signal strength is obtained by detecting the RSSI value, and the network load balancing metric is monitored from the network traffic distribution status. After the collection is completed, compare the parameter differences between different devices. For example, the response time of device D1 under node N1 is 120ms, the signal strength is -65dBm, and the load metric is 0.6, compared with the response time of 150ms, signal strength of -70dBm, and load metric of 0.8 of device D2 under the same node, to determine which device has better connection efficiency. In this process, the lower the response time, the closer the signal strength is to 0, and the smaller the load metric, the better the performance. Perform standard normalization operations on the three metrics respectively, calculate the sorting priority through the metric difference between devices, adopt the multi-round comparison method to reasonably arrange the connection request order according to the device resource utilization rate, and record the sorting sequence.

[0093] S312: Call the connection request sorting table, and according to the response time and signal strength of the network node to which the device is connected, use the formula:

[0094]

[0095] Calculate the access priority score PR, and automatically sort the connection requests according to the score to obtain the device connection sorting result. Among them, GR ij represents the response time of device i under node j, reflecting the time from when the device sends a signal to when it receives a response. GR max represents the maximum response time of all devices under all nodes, which is a reference value used to standardize the response time. GS ij represents the signal strength of device i under node j, which is a key indicator to measure the network connection quality and affects the stability and speed of data transmission. GS max represents the maximum signal strength of all devices under all nodes, which is a reference value used to standardize the signal strength. GL ij represents the network load balancing metric of device i under node j, which is used to evaluate the balance of network load. GL max represents the maximum network load balancing metric of all devices under all nodes, which is a reference value used to standardize the network load balancing metric. Npr is the weighted average total number of items, which is 3 and includes response time, signal strength, and network load balancing metrics;

[0096] Call the sorting positions of each device in the connection request sorting table, and analyze the response time, signal strength, and load balance under the current connection node again. Calculate the score according to the parameter application formula. For example, the response time of device D2 under node N1 is 150ms, GR max is set as the maximum response time of 150ms among the same batch of devices, GS ij is the signal strength value of the device under this node, with the unit of dBm, which needs to be measured through RSSI. For example, D2 is measured as -70dBm, corresponding to the maximum signal strength GS max is -60dBm. During actual processing, take its absolute value and convert it to a positive value to participate in the normalization calculation, and get GL ij is the network load index, and its value range is between 0 and 1. For example, D2 is 0.8, GL max is the maximum load value of 0.8, and after normalization, it is 1.0000. Since N pr = 3, indicating an equal-weighted average of the three parameters. Substitute into the formula to calculate the access priority score of D2:

[0097]

[0098] Similarly, if the response time of D3 under node N2 is 90ms, the signal strength is -60dBm, and the load index is 0.5, substitute into the calculation to get:

[0099]

[0100] This score represents the access priority of the device under the current node. The higher the score, the better the comprehensive performance. After sorting, the device connection sorting result is obtained, and this result is used as the basis for subsequent priority connection judgment.

[0101] S313: According to the device connection sorting result, mark the high-priority devices with the right to connect first, and postpone the connection of the remaining devices. At the same time, call the connection request sorting table to sort the postponed devices again to obtain the access priority index;

[0102] First, read the device number with the highest sorting value and mark its priority according to the preset access scoring benchmark value. Set the access scoring benchmark value to 0.8. If the scoring value is greater than this benchmark value, it is given the priority to connect. For example, the scoring value of D2 is 1.0556, which is significantly greater than the benchmark value of 0.8. Therefore, D2 is marked as a high-priority device. Then, read D3 with a scoring value of 0.7417 and D4 with a scoring value of 0.7333. Since their scores are lower than the access scoring benchmark value, they are classified as ordinary devices and do not participate in the first-round connection allocation. This process is based on the numerical comparison and judgment between the scoring results and the benchmark value. Subsequently, rewrite the connection requests of the marked ordinary devices into the connection request sorting table, sort them again according to their original sorting values, and arrange the remaining requested devices in descending order of their scoring values. For example, D3 is ranked before D4. The updated sorting table is used as the execution basis for the final connection strategy, and the sorting data is mapped to the access classification control system to generate a unified connection order control instruction, and finally obtain the access priority index.

[0103] Please refer to Figure 5 , and the specific steps for obtaining the task load management information are as follows:

[0104] S411: Based on the access priority index, evaluate the real-time task queue data and the current memory usage data of the connected devices, compare the number of task requests with the set device request capacity, and compare the memory usage with the memory capacity to judge the current load level of the devices and obtain the device load judgment result;

[0105] Obtain the real-time task queue data and current memory usage data of all connected devices. The access priority can be established based on dimensions such as device type, running time, physical location, or historical performance. For example, prioritize the access of critical data collection devices. Assume device D01 is an edge node, and its current task queue contains 25 uncompleted requests. This data can be obtained through the system polling command line or SNMP protocol. At the same time, by invoking the memory monitoring of the operating system, it can be obtained that the device has currently used 5.5GB of memory, and the total memory of the device is 8GB. Then the remaining memory capacity of the device is 2.5GB. Calculating the memory usage rate is 5.5÷8 = 0.6875, that is, 68.75%. If the system presets the task capacity threshold to 30 and the memory usage rate threshold to 70%, then it is necessary to compare the current task request quantity of the device with the task capacity threshold. If the task number 25 is less than the threshold 30, this item is judged as normal. Comparing the memory usage rate, it is found that 68.75% is also less than 70%. Since both results are lower than the preset threshold, it is judged that the device load is normal, and the corresponding load judgment value is set to 0.6875. If it is replaced with another device D02, its task request number is 40, and the memory usage is 6.2GB. Calculating the memory usage rate is 6.2÷8 = 0.775, that is, 77.5%. Both exceed the preset threshold. At this time, the task request is judged as 1.0, and the memory usage is also 1.0. The comprehensive load judgment value is 1.0. Therefore, in this process, it is necessary to perform a dual comparison operation of the actual value and the set threshold for each item of data, and use the maximum value as the final device load judgment result. This judgment value will be used as an important reference basis for task allocation adjustment in the future.

[0106] S412: Based on the device load judgment result, use the formula:

[0107]

[0108] Calculate the task allocation offset value ΔGD for adjusting tasks to optimize the device's workload. Among them, CU represents the data processing performance parameter of the device, which is a coefficient quantifying the device's processing ability and is used to evaluate the device's performance under a specific task load. QD represents the current task request number of the device, reflecting the number of tasks that the device is currently processing or waiting to process. RD represents the task response rate, which is an indicator measuring the speed or efficiency of the device's response to task requests. MD represents the free memory capacity of the device, indicating the amount of memory that the device has not been used at a specific time point. K GD is an adjustment coefficient used to adjust the task allocation ratio according to the relationship between device performance and load;

[0109] Combined with parameters such as the data processing performance parameters, number of task requests, response rate, and remaining memory capacity corresponding to each device. Taking device D01 as an example, its processing performance CU is 3.2 GFLOPS, which is recorded as 0.53 after normalization. The current number of task requests QD is 25, normalized to 0.42. The task response rate RD is 50 req / s, and the corresponding normalized value is 0.51. The idle memory MD is 2.5 GB, and the normalized value is 0.38. The adjustment coefficient K_GD takes the set value of 1.1 and remains unchanged. Substitute into the formula:

[0110]

[0111] Taking device D02 as an example, its CU is 4.5, normalized to 0.74. QD is 40, corresponding to a normalization of 0.67. RD is 65, normalized to 0.68. MD is 1.8, normalized to 0.29. K_GD is 1.1. Substitute into the formula for calculation:

[0112]

[0113] It is obtained that the task allocation offset value of device D01 is 0.2751, and that of D02 is 0.5622. This offset value will be used for setting the task weight in the subsequent task ratio regulation step. The higher the value, the greater the task allocation pressure of the current device, and its task proportion needs to be relatively reduced to balance the execution pressure among devices. Finally, the task allocation offset value is output.

[0114] S413: According to the task allocation offset value, reallocate the proportion of the device task requests, and combined with the original task queue structure, reconstruct the scheduling information field to obtain the task load management information;

[0115] Statistical settings are made for the total amount of task requests. For example, the total number of tasks to be allocated in the current system scheduling is 90. It is necessary to calculate the task allocation ratio according to the offset values of each device. The offset values are calculated using the ratio normalization method. The total offset value is 0.2751 + 0.5622 = 0.8373. The proportion that D01 should bear is 0.2751 ÷ 0.8373 = 0.3285, that is, 32.85%. For D02, it is 0.5622 ÷ 0.8373 = 0.6715, that is, 67.15%. Subsequently, the total number of tasks is allocated. The number of tasks allocated to D01 is 90 × 0.3285 = 29.565, and the result is rounded down to 29. The number of tasks allocated to D02 is 90 × 0.6715 = 60.435, and it is rounded down to 60. The remaining 1 task can be processed by the device with the lowest response time according to the minimum response time field in the task queue. Here, it is recorded as a part of the remaining supplementary processing mechanism after allocation. This adjustment strategy performs task reallocation by combining the current operating state of the device and the result of offset comparison. In the actual scenario, the task scheduling field of D01 will be updated to the task manager, including key elements such as the updated fields "device code = D01", "allocated task number = 29", "start and end range of task segment numbers = 0001 - 0029", "scheduling timestamp = 20250409T102500", "expected response window = 3 - 7 seconds", etc. Similarly, the D02 field structure will include the device code = D02, the allocated task number = 60, the task number range = 0030 - 0089, etc. The scheduling field will finally be written into the scheduling control table and encapsulated in a JSON structure, and updated synchronously on the task scheduling and control bus to further ensure the consistency of data status between the scheduling control layer and the device task layer during the task execution stage. After integrating all device scheduling information, the system generates a unified field structure and outputs task load management information.

[0116] Please refer to Figure 6 , and the specific steps for obtaining the device connection control feedback result are as follows:

[0117] Call the task load management information, and based on the current working state of the intelligent device, compare the corresponding relationship between the task execution frequency and the memory usage rate, determine whether the task load exceeds the standard, and combine the number of intelligent device threads and the task period interval to determine the task load change trend;

[0118] First, extract the task execution frequency and calculate the single-thread processing load based on the number of threads. For example, if the task frequency of a certain device is 20 Hz and the number of threads is 8, then the single-thread load is 2.5 Hz. Use this indicator to judge the task scheduling pressure distribution. Then, collect the memory usage rate and judge the running state of the current device according to the preset memory usage threshold. The threshold is divided into three intervals: 0%–70% is for controllable load, 70%–85% is for tense load, and above 85% is for overloaded load. If the current memory usage of this device is 92%, it has obviously entered the overloaded interval. At this time, analyze the task scheduling density in combination with the task cycle and connection time interval. Suppose the task cycle is 1600 ms and the connection time interval is 300 ms, then the theoretical number of connections per unit cycle is 5 times. Combining with the frequency, it can be known that the number of tasks to be processed in this cycle is 20 Hz × 1.6 s = 32 times. Distribute 32 tasks to 5 connection operations, and on average, each connection needs to execute 6.4 tasks. This density exceeds the single-connection load capacity. Combining the current memory pressure, it is determined that the task density is overloaded. On the contrary, if the task frequency is 10 Hz, the number of threads is 4, the memory usage is 65%, the cycle is 2000 ms, and the connection interval is 500 ms, then the number of connections in the cycle is 4 times, the total number of tasks is 20 times, and on average, each connection needs to process 5 tasks. And the single-thread load is 2.5 Hz. Combining with the memory state in the controllable interval, it is determined that the task density is reasonable. Through the linkage comparison and judgment of the above multiple indicators, the system evaluation of the task execution frequency, thread load, memory pressure, and connection frequency is completed. Finally, according to the task volume density, memory capacity, and time matching degree, the task density floating trend of the device under the current configuration is obtained, and the task load status in the current cycle is summarized accordingly.

[0119] Based on the task load change trend, analyze the matching degree between the task load and the connection cycle. According to the changes in the task frequency and the device connection success rate, synchronously adjust the device connection frequency and task cycle configuration to obtain the device connection control feedback result;

[0120] Combined with the current connection time interval of the device and the task cycle, further analyze the coverage of task density in the connection cycle and its dynamic adaptation ability to determine whether there are problems of insufficient connection frequency configuration or excessive redundancy. In the specific process, first map the task density change value to the connection frequency according to the time distribution for matching analysis. Suppose the current task cycle of a certain device is 1800 ms and the connection time interval is 400 ms, then the number of connections within the cycle is 4.5 times, rounded up to 4 times. If the task frequency is 15 Hz, the total number of tasks is 27 times. On average, each connection needs to process 6.75 tasks. If the current number of threads is 6, each thread needs to bear 1.125 tasks. If the memory utilization rate of this device is 80% and it is in the "tense" interval, and the connection success rate is 85%, which is lower than the benchmark threshold of 90%, it indicates that there is a downward trend in connection efficiency. At this time, it is necessary to judge whether the connection performance has decreased due to the increase in task load density. According to the condition that the current number of connections is not enough to cover the task density, it is recommended to adjust the connection time interval to 300 ms, so as to increase the number of connections per cycle to 6 times, and reduce the task pressure per connection to 4.5 times. The thread sharing intensity also decreases accordingly. At the same time, compare the connection success rate and task loss rate indicators of the device in the past three cycles. If the connection success rate increases to more than 90% and the task loss rate drops to the tolerable range after the adjustment, it can be confirmed that this adjustment is targeted. If there is no improvement, it is necessary to continue to adjust the task frequency or extend the task cycle to relieve the density pressure. Through this way, establish a dynamic feedback mechanism between the adjustment of the connection control strategy and the task density state, and obtain the device connection control feedback result.

[0121] An intelligent device remote collaborative control system based on the Internet of Things, the system includes:

[0122] The device status monitoring module is based on the Internet of Things platform, monitors the CPU utilization rate and memory utilization rate of the intelligent device, regularly collects the status data of the intelligent device, and dynamically adjusts the data transmission interval through real-time analysis of the intelligent device performance to obtain the device performance monitoring data set;

[0123] The network load management module obtains the bandwidth utilization rate, signal strength and network response time of the intelligent device in the edge network according to the device performance monitoring data set, compares the network parameters, and identifies and adjusts the network nodes with the load reaching the upper limit to obtain the network load balance index;

[0124] The device access priority module analyzes the CPU utilization rate and memory usage of the intelligent device based on the network load balance index, and automatically sorts the connection requests according to the response time and signal strength of the network nodes, marks the high-priority devices with the priority connection right, and delays the connection of the remaining devices to obtain the access priority index;

[0125] Based on the access priority metric, the intelligent task scheduling module evaluates the task queue length and memory usage rate of the connected devices, and adjusts the task allocation ratio according to the data processing performance of the devices to obtain the task load management information;

[0126] The connection dynamic regulation module calls the task load management information, analyzes the task execution density and memory usage status, dynamically adjusts the connection time interval and task period to match the real-time task requirements, and synchronously adjusts the device connection frequency to obtain the device connection control feedback result.

[0127] The above are only the preferred embodiments of the present invention, and do not limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution content of the present invention still belong to the protection scope of the technical solution of the present invention.

Claims

1. A method for remotely collaborative control of intelligent devices based on the Internet of Things, characterized in that, It includes the following steps: S1: Based on the Internet of Things platform, monitor the CPU utilization rate and memory usage rate of intelligent devices, regularly collect the status data of intelligent devices using the Internet of Things communication interface, and dynamically adjust the data transmission interval by analyzing the performance of intelligent devices in real time to obtain a device performance monitoring data set; S2: According to the device performance monitoring data set, obtain the bandwidth utilization rate, signal strength, and network response time of intelligent devices in the edge network, compare the network parameters, identify and adjust the network nodes with the load reaching the upper limit, and obtain the network load balance index; S3: Based on the network load balance index, analyze the CPU utilization rate and memory usage of intelligent devices, automatically prioritize connection requests according to the response time and signal strength of the network nodes connected to each intelligent device, mark high-priority devices with the right to priority connection, and postpone the connection of the remaining devices to obtain the access priority index; S4: Based on the access priority index, evaluate the task queue length and memory usage rate of the connected devices, and adjust the task allocation ratio according to the data processing performance of the devices to obtain the task load management information.

2. The intelligent device remote collaborative control method based on the Internet of Things according to claim 1, wherein, The device performance monitoring data set includes performance trends, resource occupancy rates, and status feedback information. The network load balance index includes load capacity and adjustment response results. The access priority index includes weight levels, sorting logics, and response priorities. The task load management information includes processing capabilities, queue optimization results, and task response speeds.

3. The method for remotely and collaboratively controlling intelligent devices based on the Internet of Things according to claim 1, wherein The specific steps for obtaining the device performance monitoring data set are as follows: S111: Based on the Internet of Things platform, use the communication interface to collect the status data of intelligent devices, monitor the CPU utilization rate, memory usage rate, and current network bandwidth occupancy, identify the missing data and fill it to obtain the resource utilization time series; S112: According to the resource utilization time series, analyze the status data of intelligent devices in consecutive operation cycles, calculate the change rate within a certain time, compare it with the known stable change interval, and mark the time periods exceeding the fluctuation benchmark to obtain the stability fluctuation trend; S113: For the stability fluctuation trend, analyze the fluctuation degree interval in the current cycle and the fluctuation change level in the historical stable cycle, and use the formula: Obtain the performance change intensity DE of the i-th device i , and dynamically adjust the data transmission interval to obtain a device performance monitoring data set, where Δce i is the CPU utilization change value of the i-th device, Δme i is the memory occupancy change value of the i-th device, be i is the bandwidth fluctuation frequency of the i-th device, re i is the response time lag factor of the i-th device, de i is the data interaction density factor of the i-th device, qe ik is the concurrent device impact coefficient of the i-th device in the k-th time period, n de is the total number of time periods.

4. The method for remotely and collaboratively controlling intelligent devices based on the Internet of Things according to claim 1, wherein The specific steps for obtaining the network load balance index are as follows: S211: According to the device performance monitoring data set, obtain the corresponding bandwidth utilization rate, signal strength, and network response time of intelligent devices in the edge network, align the time series of network parameters, and group them according to the network node positions to establish the initial network load data; S212: Call the initial network load data, compare the load of each network node with the average load of all nodes, and use the formula: Calculate the network resource transfer performance RT j , identify and adjust the network nodes with the load reaching the upper limit to obtain the network load balance index, where BT j represents the bandwidth utilization rate of the current node j, and BT z represents the bandwidth utilization rate of other nodes z compared with the current node j, ST z represents the signal strength of other nodes z, and TR z represents the network response time of other nodes z. α and β are the weights for adjusting the proportion of the bandwidth difference and the signal strength to the response time, and n rt is the total number of nodes participating in the calculation.

5. The method for remotely and collaboratively controlling intelligent devices based on the Internet of Things according to claim 1, characterized in that, The specific steps for obtaining the access priority index are as follows: S311: Based on the network load balance index, analyze the CPU utilization rate and memory usage of intelligent devices, collect the network node data of each intelligent device, sort the connection requests of different devices, and obtain the connection request sorting table; S312: Invoke the connection request sorting table, and according to the response time and signal strength of the network nodes connected by the device, use the formula: Calculate the access priority score PR, and automatically sort the connection requests according to the score to obtain the device connection sorting result, where GR ij represents the response time of device i under node j, GR max represents the maximum response time of all devices under all nodes, GS ij represents the signal strength of device i under node j, GS max represents the maximum signal strength of all devices under all nodes, GL ij represents the network load balance index of device i under node j, GL max represents the maximum network load balance index of all devices under all nodes, N pr is the weighted average total number; S313: According to the device connection sorting result, mark the high-priority devices with the right of priority connection, postpone the connection of the remaining devices, and at the same time, invoke the connection request sorting table to sort the postponed devices again to obtain the access priority index.

6. The method for remotely and collaboratively controlling intelligent devices based on the Internet of Things according to claim 1, wherein The specific steps for obtaining the task load management information are as follows: S411: Based on the access priority index, evaluate the real-time task queue data and the current memory usage data of the connected devices, compare the number of task requests with the set device request capacity, and compare the memory usage with the memory capacity to judge the current load level of the devices and obtain the device load judgment result; S412: Based on the device load judgment result, use the formula: Calculate the computing task allocation offset value ΔGD, where CU represents the data processing performance parameter of the device, QD represents the current task request number of the device, RD represents the task response rate, MD represents the free memory capacity of the device, and K GD is an adjustment coefficient; S413: According to the task allocation offset value, perform proportional reallocation on the device task request numbers, and combine the original task queue structure to reconstruct the scheduling information field to obtain the task load management information.

7. The method for remotely and collaboratively controlling intelligent devices based on the Internet of Things according to claim 1, wherein The steps further include: S5: Invoke the task load management information, analyze the task execution density and memory usage status according to the current working state of the intelligent device, dynamically adjust the connection time interval and task period, match the real-time task requirements, and synchronously adjust the device connection frequency to obtain the device connection control feedback result; The device connection control feedback result includes connection adjustment data, cycle update information, and frequency synchronization index.

8. The method for remotely and collaboratively controlling intelligent devices based on the Internet of Things according to claim 7, wherein The specific steps for obtaining the device connection control feedback result are as follows: Invoke the task load management information, according to the current working state of the intelligent device, compare the corresponding relationship between the task execution frequency and the memory usage rate, judge whether the task load exceeds the standard, and combine the number of intelligent device threads and the task cycle interval to determine the task load change trend; Based on the task load change trend, analyze the matching degree between the task load and the connection cycle, and synchronously adjust the device connection frequency and task cycle configuration according to the changes in the task frequency and the device connection success rate to obtain the device connection control feedback result.

9. An intelligent device remote collaborative control system based on the Internet of Things, characterized in that, The system is used to implement the method for remotely coordinating the control of intelligent devices based on the Internet of Things according to any one of claims 1-8. The system includes: The device status monitoring module, based on the Internet of Things platform, monitors the CPU utilization rate and memory usage rate of the intelligent device, regularly collects the status data of the intelligent device, and dynamically adjusts the data transmission interval through real-time analysis of the intelligent device performance to obtain the device performance monitoring data set; The network load management module, according to the device performance monitoring data set, obtains the bandwidth utilization rate, signal strength, and network response time of the intelligent device in the edge network, conducts data comparison on the network parameters, identifies and adjusts the network nodes with the load reaching the upper limit to obtain the network load balance index; The device access priority module, based on the network load balance index, analyzes the CPU utilization rate and memory usage of the intelligent device, automatically sorts the connection requests according to the response time and signal strength of the network nodes, marks the high-priority devices with the right of priority connection, and postpones the connection of the remaining devices to obtain the access priority index; Based on the access priority metrics, the intelligent task scheduling module evaluates the task queue length and memory usage rate of the connected devices, adjusts the task allocation ratio according to the data processing performance of the devices, and obtains task load management information; The connection dynamic regulation module calls the task load management information, analyzes the task execution density and memory usage status, dynamically adjusts the connection time interval and task period, matches the real-time task requirements, and synchronously adjusts the device connection frequency to obtain the device connection control feedback result.

Images