A dynamic collaborative management method for industrial wireless networks for serial monitoring tasks

Abstract

The present invention discloses a method for dynamic collaborative management of industrial wireless networks for serial monitoring tasks, and relates to the field of industrial Internet of Things. A dynamic priority index is introduced that simultaneously considers the number of times a sensor is called and the average monitoring error of the sensor. Based on this dynamic priority index, network resources such as wireless channels are allocated by taking into account two time scales, namely, the total monitoring task and the monitoring subtask. The present invention overcomes the defects of network resource management technology in industrial wireless monitoring scenarios involving multiple serial subtasks, reduces the energy consumption of wireless sensors while ensuring the monitoring effect, and increases the life of wireless sensors.

Description

Technical Field

[0001] The present invention relates to the field of industrial Internet of Things, and in particular to an industrial wireless network dynamic collaborative management method oriented to serial monitoring tasks. Background Art

[0002] In recent years, with the continuous development of wireless communication technology, the Industrial Internet of Things (IIoT) has also been more widely used in the field of intelligent manufacturing. In industrial sites, a large number of wireless sensors are used for remote monitoring, such as remote measurement of multi-active surface angles during aircraft assembly and testing and remote temperature detection of steel plates during hot rolling. In the actual production process, a total monitoring task will contain multiple monitoring subtasks, and each monitoring subtask has equal importance. These subtasks are usually in a serial relationship and will be executed one by one in a certain order. The monitoring scope involved in each monitoring subtask is different, and the sensors that need to be scheduled also change accordingly. Therefore, the total number of times different sensors are called in a total detection task is different, which will cause some sensors to need to participate in multiple monitoring subtasks more times during the total detection task. At the same time, since the energy of each wireless sensor is limited, if too much energy is consumed in the previous subtask, it will affect the life of the sensor and may also affect the monitoring quality of subsequent monitoring subtasks. In addition, most wireless sensors are powered by batteries, and the position of each measuring point is fixed during operation, and the position of the sensor cannot be adjusted frequently. According to this actual situation, when allocating network resources during the transmission process, it is necessary to give priority to allocating resources to sensors that are scheduled more times in a total monitoring task, so as to improve the transmission success rate of the sensor and ensure the monitoring quality.

[0003] In addition, compared with the outdoor environment, although the industrial site does not need to consider the influence of weather, the movement of workers, the position of the robotic arm, and the trajectory of the automatic tracking robot will show periodic or non-periodic changes over time. The occlusion caused by these factors will affect the quality of the wireless channel, causing the channel state to dynamically change, affecting the transmission quality, and causing packet loss and other phenomena. In the monitoring subtask, the wireless sensor needs to transmit the monitoring data to the edge server through the wireless network for analysis. The packet loss caused by the unsatisfactory communication environment will affect the monitoring quality. Specifically, continuous packet loss will cause the edge device to be unable to accurately capture important parameters of the production process, resulting in greater monitoring errors and reduced monitoring quality. To this end, it is necessary to dynamically allocate wireless network resources within each monitoring subtask according to the actual transmission situation and monitoring error to ensure the monitoring quality of the current monitoring subtask.

[0004] In view of the different requirements on the two time scales of the total monitoring task and the monitoring subtask, it is necessary to design a reasonable network resource management strategy so that both are taken into consideration and coordination between multiple serial monitoring subtasks is achieved. From the perspective of the total monitoring task, sensors that are called more times can reduce power consumption while ensuring the completion of the monitoring task, extend the working time as much as possible, and ensure the monitoring quality in each monitoring subtask. From the perspective of the monitoring subtask, the allocation of network resources needs to consider the actual transmission situation and monitoring errors for dynamic adjustment to improve the monitoring quality of the monitoring subtask. How to take into account the two time scales of the total monitoring task and any monitoring subtask, allocate network resources more efficiently, coordinate between multiple monitoring subtasks at the overall level, and ensure monitoring quality at the level of a single subtask is an urgent problem to be solved.

[0005] After searching the existing literature, it is found that the most similar implementation scheme is the Chinese patent application number: 201610327027.0, named: A priority-based multi-service heterogeneous wireless network resource allocation method. The specific approach is: when multiple mobile terminal services arrive at the same time, the use of Sigmoid-form utility function can easily unify the modeling of multiple types of services and assign priorities to different services. Based on the result of the priority allocation, each terminal is connected to a suitable wireless network to meet the business needs of each mobile terminal. Although this method sets the business priority when allocating resources, the industrial field environment is complex. It is difficult to accurately characterize the importance of different sensors by calculating the priority only according to the Sigmoid utility function for resource allocation. The proposed method cannot be applied to industrial wireless monitoring systems.

[0006] In terms of research papers, multi-sensor wireless monitoring systems have also received some attention. These include "Transmission scheduling for multi-process multi-sensor remote estimation via approximate dynamic programming", "Multi-sensor transmission power control for remote estimation through a SINR-based communication channel", "Optimal Scheduling of Multiple Sensors Over Lossy and Bandwidth Limited Channels", etc. However, the solutions proposed in these papers have two main shortcomings: 1. Such studies consider the network resource allocation problem within a single monitoring subtask, without considering the scenario where a total monitoring task contains multiple serial monitoring subtasks, and without considering the actual situation that the total number of times different sensors are called is different, making it difficult to achieve coordination of network resource allocation between multiple monitoring subtasks. 2. Most of these consider the scenario where multiple sensors share a single wireless channel. Since multiple users share a channel, the same-frequency interference will cause the sensor to use a larger transmission power consumption to overcome the interference, which will affect the life of the sensor. If there are subsequent monitoring subtasks, it will also affect the monitoring quality of the subsequent subtasks. Therefore, a wireless resource allocation method that is more suitable for industrial sites is needed.

[0007] Existing network resource management methods rarely take into account the actual situation that the total monitoring task of the industrial wireless monitoring system contains multiple serial monitoring subtasks, and different sensors are scheduled at different times in a total monitoring task. The allocation of network resources is not considered from the time scale of the total monitoring task, which may cause the energy consumption of the preceding subtasks of some sensors to be too high, affecting the life of the sensors and further affecting the monitoring quality of subsequent monitoring subtasks.

[0008] Some existing network resource management methods in multi-sensor remote monitoring scenarios simply rely on the cumulative number of packet losses or monitoring errors as dynamic priority indicators. Such methods can only manage network resources within a single monitoring subtask, and are not suitable for monitoring tasks consisting of multiple serial monitoring subtasks, and it is difficult to achieve coordination between multiple monitoring subtasks.

[0009] The existing network resource management method in the industrial wireless monitoring scenario does not consider the impact of the number of sensor calls in multiple monitoring subtasks and the unstable wireless network in a single monitoring subtask on the monitoring quality. It is unable to dynamically adjust the importance of any sensor in the monitoring subtask, which will affect the management quality of wireless network resources.

[0010] Therefore, technicians in this field are committed to developing a dynamic collaborative management method for industrial wireless networks for serial monitoring tasks. To overcome the defects of network resource management technology in industrial wireless monitoring scenarios containing multiple serial subtasks, a dynamic priority index that considers both the number of sensor calls and the average monitoring error of the sensor is introduced. It can take into account the two time scales of the total monitoring task and the monitoring subtask, and allocate network resources such as wireless channels. While ensuring the monitoring effect, the energy consumption of wireless sensors is reduced and the life of wireless sensors is increased. Summary of the invention

[0011] In view of the above-mentioned defects of the prior art, the technical problem to be solved by the present invention is how to initialize the priority of different wireless sensors according to the calling status of each sensor in multiple serial monitoring subtasks within a total monitoring task; how to form a dynamic priority index based on the actual situation and monitoring quality of the data sent by any current wireless sensor in any monitoring subtask; and how to design an effective network resource allocation scheme based on the dynamic priority index to ensure the transmission quality of sensors with higher dynamic priorities, while improving the monitoring quality, reducing the average energy consumption of the sensor in the entire detection task process, and realizing the coordination between multiple serial monitoring subtasks.

[0012] To achieve the above object, the present invention provides an industrial wireless network dynamic collaborative management method for serial monitoring tasks, comprising the following steps:

[0013] Step 1: Count the number of times all sensors are called in all monitoring subtasks of a total monitoring task and arrange them from high to low according to the numerical value; take the largest number of times called as the benchmark, normalize the number of times other sensors are called as the initial priority value of all sensors;

[0014] Step 2: In any monitoring subtask, an observation window is set, which includes multiple transmission cycles, each of which contains a limited number of time slots for retransmission. In each observation window, the packet loss rate of each sensor is counted according to the actual transmission result, and the average monitoring error index is calculated according to the system model. Similar to the normalization process in step 1, the average monitoring error of all sensors is normalized.

[0015] Step 3: According to the set weight, the initial value of the priority and the dynamic average monitoring error are integrated to obtain the dynamic priority index;

[0016] Step 4: Estimate the average channel gain of a future observation window period based on the historical channel state information in the observation window period and the channel state information estimation algorithm, and sort all available channels from high to low according to the channel gain;

[0017] Step 5: In the first observation window of each monitoring subtask, wireless channels are allocated according to the initial priority values ​​of the called sensors, and channels with high channel gains are preferentially allocated to sensors with high initial priorities for sending monitoring data;

[0018] After the first observation window period ends, the wireless channels are allocated according to the dynamic priorities obtained in step 2 and step 3, and channels with good channel status are preferentially allocated to sensors with high dynamic priorities for sending monitoring data;

[0019] Step 6: The sensor sends data in multiple time slots within a transmission cycle according to the allocated wireless channel, and determines whether the data is successfully transmitted based on the ACK signal to decide whether to retransmit; before each monitoring subtask is completed, return to step 2 until the next monitoring subtask starts.

[0020] Furthermore, in step 1, for all monitoring subtasks of the overall monitoring task, the number of times each sensor is called in all monitoring subtasks is counted, and based on the statistical values, the initial priority values ​​of the sensors are set and normalized from high to low, and the priorities of the sensors are divided based on the time scale of the overall monitoring task. In the subsequent network resource allocation process, reference indicators are preferentially provided to sensors that are called more times, so as to achieve coordination between multiple monitoring subtasks.

[0021] Furthermore, in step 2, during any monitoring subtask, an observation window is set, and the edge server can calculate the average monitoring error of each sensor through the packet loss situation of the sensor and the monitoring system model; by setting the observation window, the cost of communication resource allocation is reduced while improving the dynamic resource allocation capability.

[0022] Furthermore, in step 2, according to the actual transmission situation of each sensor, the edge server counts the packet loss rate of each sensor, and calculates the monitoring error caused by a single transmission according to the measurement error of each sensor and the system model, which is recorded as the average monitoring error.

[0023] Furthermore, in step 2, the higher the packet loss rate of the sensor and the higher the sensor measurement error, the greater the average monitoring error.

[0024] Furthermore, in step 3, the calculation method of the dynamic priority index of the sensor is to set weights to perform weighted fusion of the initial priority value of each sensor and the average monitoring error, and the weights of the initial priority value and the average monitoring error are both positive numbers and the sum is 1; from the time scale of the total monitoring task, the index based on the number of times the sensor is called is considered, and on the time scale of a monitoring subtask, the average monitoring error index due to transmission packet loss and sensor measurement error is considered, and the two indicators are normalized and weighted fused.

[0025] Furthermore, in step 3, different fusion weights reflect the importance between the overall monitoring task and the individual monitoring subtasks.

[0026] Furthermore, in step 4, the channel state information estimation algorithm estimates the available channel state of a future observation window based on the historical channel state information after the first observation window of any monitoring subtask.

[0027] Furthermore, in step 5, based on the dynamic priority indicator, wireless channels are allocated. In the first observation window of each monitoring subtask, since the average monitoring error of all sensors is 0, the edge server allocates wireless channels according to the initial priority value of the sensor in the monitoring subtask, and allocates the channel with the best status among the available wireless channels to the sensor with the highest initial priority value, thereby ensuring that sensors that are called more times will be preferentially allocated wireless channels with higher channel gains.

[0028] Furthermore, in step 5, based on the dynamic priority indicator, wireless channels are allocated. At the beginning of the second observation window of each monitoring subtask, the edge server allocates wireless channels according to the dynamic priority value of the sensor in the task, and allocates the best channel in the available wireless channels to the sensor with the highest dynamic priority, thereby improving the transmission success rate and reducing the number of retransmissions and transmission energy consumption.

[0029] In a preferred embodiment of the present invention, the purpose of the present invention is to overcome the defects of network resource management technology in industrial wireless monitoring scenarios containing multiple serial subtasks, and introduce a dynamic priority index that simultaneously considers the number of times the sensor is called and the average monitoring error of the sensor. Based on this dynamic priority index, the two time scales of the total monitoring task and the monitoring subtask can be taken into account to allocate network resources such as wireless channels. While ensuring the monitoring effect, the energy consumption of wireless sensors is reduced and the life of wireless sensors is increased.

[0030] To achieve the above object, the present invention adopts the following technical solutions:

[0031] The industrial wireless monitoring system as a whole consists of an edge server, multiple wireless monitoring sensors, an industrial wireless network, and multiple monitored systems or devices. The industrial wireless monitoring task consists of a total monitoring task and multiple monitoring subtasks. We record the time scale of a total monitoring task as a monitoring work cycle. Each monitoring subtask contains multiple transmission cycles, and each transmission cycle contains a limited number of time slots. The sensor will sample according to a certain sampling cycle, and then send it to the edge server through the wireless network for analysis and processing.

[0032] Step 1: Set the initial priority values ​​of all sensors.

[0033] 1.1 According to all monitoring subtasks within a pre-set monitoring task, count the number of times each sensor in all monitoring subtasks is scheduled and sort them from high to low.

[0034] 1.2 The number of times all sensors are called is normalized according to the maximum value, and the corresponding initial priority value is set to 1. The initial priority value of other sensors is the ratio of the number of times the sensor is called to the maximum value. At this point, the normalized initial priority value of multiple sensors based on the number of times called can be obtained, and this initial priority value is stored in each sensor and edge server.

[0035] Step 2: Update the average monitoring error of sensors within the monitoring subtask.

[0036] In order to realize the dynamic allocation of network resources within the monitoring subtask, an observation window is set up, which contains a finite number of transmission cycles, and each transmission cycle contains a finite number of time slots. The edge server calculates and updates the average monitoring error of each sensor based on the transmission results of all transmission cycles within this observation period. In each monitoring subtask, there are some observation windows. If the last observation window is less than a finite transmission cycle, the actual situation shall prevail. However, this will not affect the execution of the algorithm, because the result of the last observation window will not affect the next monitoring subtask.

[0037] 2.1 During the execution of each monitoring subtask, for each sensor, the edge server can judge the result of each transmission based on whether the sensor data is received in each time slot. Due to the instability of the industrial wireless channel state, sensor data may be lost. Record the cumulative number of packet losses for each sensor in an observation period. After an observation period ends, calculate the packet loss rate of each sensor in the observation period. Then, based on the packet loss rate and the system model, calculate the average monitoring error caused by packet loss to characterize the size of the monitoring error that may be caused by the transmission of each time slot.

[0038] 2.2 In order to ensure the accuracy of monitoring, the larger the monitoring error, the higher the dynamic priority index of the sensor. We normalize the estimated errors of the remaining sensors according to the average monitoring error of the largest sensor. That is, the dynamic priority of the sensor with the largest monitoring error is 1, and the normalized average monitoring error of other sensors is the ratio of the average monitoring error of the sensor to the maximum average monitoring error. When any observation window ends, the average monitoring error of all sensors will return to zero.

[0039] Step 3: Generate dynamic priority indicators.

[0040] After each observation window, the edge server sets weights based on the initial priority value obtained in the first step and the average monitoring error obtained in the second step, and performs weighted fusion on the normalized initial priority value and the average monitoring error to form a dynamic priority index, which is used to characterize the importance of each sensor in each observation window. The weights of the initial priority value and the average monitoring error are both positive numbers and their sum is 1.

[0041] Specifically, the coefficient of the initial priority value represents the initial priority value set from the time scale of the total monitoring task, while the coefficient of the dynamic priority represents the dynamic priority of the sensor in the current monitoring subtask.

[0042] Step 4: Get and sort the wireless channel status.

[0043] The edge server has the function of an edge gateway and can capture the available channels in the current communication environment. Before each observation window, the edge server can estimate the average channel state in the next window period based on the historical information of the channel and related algorithms, and arrange the estimated average channel states from high to low to obtain a descending list of available wireless channels.

[0044] Step 5: Allocate wireless channels based on dynamic priorities.

[0045] The edge server matches the optimal wireless channel for all sensors participating in the current monitoring subtask in order of dynamic priority based on the matching principle.

[0046] 5.1 Since the initial value of the dynamic priority in each monitoring subtask is 0 and will only change after an observation period, all transmission cycles in the first observation period actually allocate wireless channels according to the initial priority value of each sensor. Specifically, in the first observation period of each monitoring subtask, the edge server sorts the sensors in descending order according to the initial priority value of the sensors in the current subtask.

[0047] 5.2 After sorting, the edge server assigns the best channel in the available wireless channels to the sensor with the highest initial priority value. Then, the assigned channels and sensors are removed from their respective sets. If there are still available channels in the channel list, the edge server continues to assign the best channel in the remaining available wireless channels to the sensor with the highest initial priority value among the remaining sensors until there are no available channels in the channel list or all sensors have been assigned wireless channels.

[0048] 5.3 After the first observation period, before each new observation period begins, the edge server will allocate wireless channels based on the dynamic priority. Specifically, according to the method for calculating the dynamic priority in the second step, the edge server calculates the dynamic priority of each sensor. Then, according to the method in the third step, the dynamic priority is updated, and the updated dynamic priority of all sensors in the monitoring subtask is obtained.

[0049] 5.4 The edge server sorts the sensors according to the dynamic priority of the sensors in the current subtask. Afterwards, similar to the approach in 5.2, the edge server allocates the best channel in the available wireless channels to the sensor with the highest dynamic priority. Then, the allocated channels and sensors are removed from their respective sets. If there are still available channels in the channel list, the edge server continues to allocate the best channel in the remaining available wireless channels to the one with the highest dynamic priority among the remaining sensors until there are no available channels in the channel list or all sensors have been allocated wireless channels.

[0050] Step 6: The sensor sends data according to the assigned wireless channel

[0051] In each transmission cycle, the sensor can send data according to the allocated wireless channel in each time slot. In order to facilitate the application of industrial sites, all sensors perform each data transmission according to the same pre-set transmission power. In each time slot, the sensor can know the result of its current transmission according to the Acknowledge Signal (ACK) signal of the transmission process. If the transmission fails, the data will continue to be sent. If the transmission is successful, the transmission of this monitoring data will end. If all transmissions in the entire transmission cycle fail, the current data packet will be discarded.

[0052] Compared with the prior art, the present invention has the following obvious substantial features and significant advantages:

[0053] 1. Considering the industrial wireless monitoring scenario that includes multiple serial monitoring subtasks, and considering the actual situation that different sensors have different calling situations in different monitoring subtasks within a monitoring task. For different wireless sensors, the priority is initialized according to the number of times they are called within a monitoring task, providing a key reference indicator on the time scale of the total detection task for the subsequent allocation of wireless network resources.

[0054] 2. In order to achieve reliable wireless channel allocation, an observation window is introduced as a channel resource allocation period, which improves the dynamic allocation capability while reducing the cost of communication resource allocation.

[0055] 3. Based on the initialization priority, the actual transmission situation and monitoring error are taken into consideration within each monitoring subtask, and the sensor average monitoring error index is introduced. It is fused with the initial priority value according to a certain weight to form a dynamic priority index, which can reflect the real importance of each sensor in each observation cycle. At the same time, the importance between the total monitoring task and a single monitoring subtask can be determined by adjusting the fusion weight when allocating network resources.

[0056] 4. In the first observation window of each monitoring subtask, wireless channels are allocated according to the initial value of the priority, ensuring that sensors with high working frequency will be given priority on a longer time scale when allocating network resources, so that sensors that are called more times can save transmission power consumption in the first observation window.

[0057] 5. After the first observation window of each monitoring subtask ends, the wireless channels are reallocated based on the dynamic priority of each sensor. The impact of monitoring error on monitoring quality is considered within a single monitoring subtask, so that sensors with high average monitoring errors will be allocated wireless channels with better channel status in the next observation window than in the previous observation window.

[0058] 6. The wireless network resource allocation method based on dynamic priority can take into account the needs of a monitoring work cycle on a large time scale and the monitoring subtask on a small time scale. Specifically, in a single monitoring subtask, the organic integration of monitoring and transmission can be achieved, and the coordination between multiple monitoring subtasks can be achieved in the overall monitoring task.

[0059] The concept, specific structure and technical effects of the present invention will be further described below in conjunction with the accompanying drawings to fully understand the purpose, characteristics and effects of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is a relationship diagram among the overall detection task, monitoring subtask and observation window of a preferred embodiment of the present invention;

[0061] Figure 2 It is a schematic diagram of an industrial wireless monitoring system of a preferred embodiment of the present invention;

[0062] Figure 3 It is a general flow chart of a method for dynamic collaborative management of industrial network resources according to a preferred embodiment of the present invention;

[0063] Figure 4 is a priority initialization flow chart of a preferred embodiment of the present invention;

[0064] Figure 5 is a dynamic priority calculation flow chart of a preferred embodiment of the present invention;

[0065] Figure 6 It is a wireless channel allocation flow chart of a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0066] The following describes several preferred embodiments of the present invention with reference to the drawings in the specification, so that the technical content is clearer and easier to understand. The present invention can be embodied in many different forms of embodiments, and the protection scope of the present invention is not limited to the embodiments mentioned in the text.

[0067] In the drawings, components with the same structure are indicated by the same numerical reference numerals, and components with similar structures or functions are indicated by similar numerical reference numerals. The size and thickness of each component shown in the drawings are arbitrarily shown, and the present invention does not limit the size and thickness of each component. In order to make the illustration clearer, the thickness of the components is appropriately exaggerated in some places in the drawings.

[0068] The present invention relates to the field of industrial Internet of Things, and relates to a method for dynamic collaborative network resource management in an industrial wireless detection scenario, in which a total monitoring task includes multiple serial monitoring subtasks. The method comprises the following steps:

[0069] Step 1: Count the number of times all sensors are called in all monitoring subtasks of a monitoring task, and sort them from high to low according to the value. Take the largest number of calls as the benchmark, and normalize the number of calls of other sensors, which is used as the initial priority value of all sensors.

[0070] Step 2: During any monitoring subtask, an observation window is set. The observation window contains multiple transmission cycles, and each transmission cycle contains a limited number of time slots for retransmission. In each observation window, the packet loss rate of each sensor is counted according to the actual transmission results, and the average monitoring error index is calculated according to the system model. Similarly, similar to the normalization process in the first step, the average monitoring error of all sensors is normalized.

[0071] Step 3: According to the set weights, the initial value of the priority and the dynamic average monitoring error are fused to obtain a dynamic priority index. Different fusion weights represent the importance of the total monitoring task and a single monitoring subtask on two time scales.

[0072] Step 4: Estimate the average channel gain in a future observation window period based on the historical channel state information in the past observation window period and the channel state information estimation algorithm, and sort all available channels from high to low according to the channel gain.

[0073] Step 5: During the first observation window of each monitoring subtask, wireless channels are allocated according to the initial priority values ​​of the called sensors, and channels with high channel gains are preferentially allocated to sensors with high initial priorities for sending monitoring data.

[0074] After the first observation window ends, the wireless channels are allocated according to the dynamic priorities obtained in the second and third steps, and the channels with good channel status are preferentially allocated to sensors with high dynamic priorities for sending monitoring data.

[0075] Step 6: The sensor sends data in multiple time slots within a transmission cycle according to the allocated wireless channel, and determines whether the data is successfully transmitted based on the ACK signal to decide whether to retransmit. Before each monitoring subtask is completed, return to step 2 until the next monitoring subtask begins.

[0076] The priority initialization method in the first step takes into account all monitoring subtasks of the total monitoring task, and counts the number of times each sensor is called in all monitoring subtasks. Based on this value, the initial priority value of the sensor is set and normalized from high to low. It can divide the priority of the sensor based on the time scale of the total monitoring task, which helps to provide reference indicators for giving priority to sensors that are called more times in the subsequent network resource allocation process, and realize the coordination between multiple monitoring subtasks.

[0077] The calculation method of the average monitoring error index of the sensor in the second step is to set an observation window during any monitoring subtask. The edge server can calculate the average monitoring error of each sensor through the packet loss situation of the sensor and the monitoring system model. By setting the observation window, the cost of communication resource allocation is reduced while improving the dynamic resource allocation capability.

[0078] The calculation method of the sensor average monitoring error index in the second step is based on the actual transmission situation of each sensor. The edge server counts the packet loss rate of each sensor, and calculates the monitoring error caused by a single transmission based on the measurement error of each sensor and the system model, which is recorded as the average monitoring error. The higher the packet loss rate and the higher the sensor measurement error, the greater the average monitoring error.

[0079] The calculation method of the dynamic priority index of sensor s in the third step Λ s =μα s +νβ s The indicator based on the number of times the sensor is called is considered from the time scale of the total monitoring task, and the average monitoring error indicator due to transmission packet loss and sensor measurement error is considered on the time scale of a monitoring subtask, and the two indicators are normalized and weighted fused. Different fusion weights can reflect the importance between the total monitoring task and a single monitoring subtask.

[0080] The channel state information estimation method in the fourth step estimates the available channel state of a future observation window based on historical channel state information after the first observation window of any monitoring subtask.

[0081] In the fifth step, the wireless channel allocation method based on dynamic priority indicators is used. During the first observation window of each monitoring subtask, since the average monitoring error of all sensors is 0, the edge server allocates wireless channels according to the initial priority value of the sensor in the subtask, and allocates the channel with the best status among the available wireless channels to the sensor with the highest initial priority value, ensuring that sensors that are called more times will be preferentially allocated wireless channels with higher channel gains.

[0082] The fifth step of the wireless channel allocation method based on dynamic priority indicators starts at the second observation window of each monitoring subtask. The edge server allocates wireless channels according to the dynamic priority values ​​of the sensors in the task, and allocates the best channel in the available wireless channels to the sensor with the highest dynamic priority, thereby improving its transmission success rate and reducing the number of retransmissions and transmission energy consumption.

[0083] Figure 1 It is a relationship diagram among the overall detection task, monitoring subtask, and observation window. The specific features are as follows:

[0084] In a monitoring task, there are multiple monitoring subtasks, which are executed one by one according to the pre-set monitoring plan, that is, each monitoring subtask is executed after the previous subtask is completed. (Take the flight control system test in the aircraft assembly process as an example. This test needs to monitor whether each aircraft active surface can reach the angle required by the flight pilot. The entire test includes 15 flight control system tests, and these 15 flight control system tests are carried out in a pre-set order.) Since the functions of the monitoring subtasks are different, the sensors required to be called for each subtask are also different. (The total number and type of active surfaces involved in each test in the flight control system test are different.) In order to achieve dynamic and collaborative network resource management within a monitoring subtask, multiple observation windows are set, and each observation window includes T transmission cycles.

[0085] Figure 2 This is a schematic diagram of an industrial wireless monitoring system. The specific features are as follows:

[0086] The entire industrial wireless monitoring system has two layers, including 1 edge server and S wireless sensors. Each sensor monitors a dynamic system or device (each active surface in the flight control system test has a contact wireless measurement sensor based on a gravity acceleration sensor, which sends the data to the edge server through the wireless network after sampling). In any monitoring subtask, the called sensor among the S sensors, after completing a sampling according to the sampling period, sends the value to the edge server through the shared wireless network for subsequent analysis and processing. There are multiple sampling periods in each monitoring subtask, and the data is sent immediately after each sampling. The edge server also has the function of an edge gateway and can manage wireless network resources in an area.

[0087] Figure 3 This is the overall flow chart of the dynamic collaborative management method of industrial networks. The specific steps are as follows:

[0088] Step 1: The edge server counts the number of times each sensor is called in all monitoring subtasks according to the arrangement of the overall monitoring task, and sets the initial priority values ​​of all S sensors.

[0089] Step 2: During the execution of any monitoring subtask, after each observation window, the edge server updates the average monitoring error of each sensor in the monitoring subtask based on the actual transmission results and the system model.

[0090] Step 3: According to the preset ratio, the initial priority value and the average monitoring error are fused to calculate the dynamic priority of each sensor in the next observation window.

[0091] Step 4: The edge server obtains the wireless channel status information in the next transmission cycle based on historical data and related algorithms, and sorts all available wireless channels from high to low according to the quality of the channel status.

[0092] Step 5: The edge server allocates a wireless channel for the sensor in the experiment to transmit data at the next moment according to the dynamic priority value of the sensor, and notifies all sensors in the area of ​​the result by broadcasting.

[0093] Step 6: Each sensor sends data according to the wireless channel assigned to it.

[0094] Figure 4 It is a priority initialization flow chart, the specific steps are as follows:

[0095] Step 1: Obtain the information of all monitoring subtasks in the overall monitoring task, that is, the number of monitoring subtasks and the number of sensors that need to be called for each monitoring subtask.

[0096] Step 2: Count the number of times all wireless monitoring sensors are called in all monitoring subtasks.

[0097] Step 3: Arrange the sensors from high to low according to the number of times they are called, take the largest value as the standard of "1", and normalize all the values.

[0098] Step 4: Use the normalized value as the initial priority value of each sensor.

[0099] Step 5: Store the initial priority value of each sensor in the edge server for later use.

[0100] Figure 5 It is a dynamic priority calculation flow chart. The specific steps are as follows:

[0101] Step 1: During each monitoring subtask, calculate the packet loss rate of each sensor based on the transmission results counted in the previous observation window.

[0102] Step 2: Based on the packet loss rate of each sensor With the monitoring system model, the average monitoring error of each sensor is calculated.

[0103] Step 3: Based on the largest average monitoring error, normalize the average monitoring errors of other sensors.

[0104] Step 4: Update the normalized value to the average monitoring error of the sensor.

[0105] Step 5: Calculate the dynamic priority by integrating the initial priority value and the average monitoring error in proportion. The coefficient of the initial priority value in the formula represents the importance of the sensor in the entire monitoring task, and the coefficient of the average monitoring error represents the importance of the sensor in the current monitoring subtask.

[0106] Figure 6 It is a wireless channel allocation flow chart, the specific steps are as follows:

[0107] Step 1: According to historical data and related algorithms, the wireless channel status information in the next transmission cycle is obtained, and all available wireless channels are sorted from high to low according to the quality of the channel status.

[0108] Step 2: Arrange all sensors of the current monitoring subtask from high to low according to the dynamic priority values.

[0109] Step 3: Assign the best channel to the sensor with the highest dynamic priority value.

[0110] Step 4: Remove the assigned wireless channels and sensors from the list.

[0111] Step 5: Check whether there are any available wireless channels remaining. If yes, return to step 3; if no, end the allocation.

[0112] The preferred specific embodiments of the present invention are described in detail above. It should be understood that ordinary technicians in the field can make many modifications and changes based on the concept of the present invention without creative work. Therefore, all technical solutions that can be obtained by technicians in the technical field based on the concept of the present invention through logical analysis, reasoning or limited experiments on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A method for dynamic collaborative management of industrial wireless networks for serial monitoring tasks, characterized in that: The following steps are involved: Step 1: Count the number of times all sensors are called in all monitoring subtasks of a monitoring task, and arrange them from high to low according to the value; Taking the maximum number of calls as the benchmark, the number of calls of other sensors is normalized as the initial priority value of all sensors; Step 2: In any monitoring subtask, an observation window is set. The observation window includes multiple transmission cycles, and each transmission cycle contains a limited number of time slots for sensor retransmission. In each observation window, the packet loss rate of each sensor is counted according to the actual transmission results, and the average monitoring error index of the sensor is calculated according to the system monitoring model. The same normalization process as in step 1 is used to normalize the average monitoring errors of all sensors; Step 3: According to the set weight, the initial value of the priority and the dynamic average monitoring error are integrated to obtain the dynamic priority index; Step 4: Estimate the average channel gain of a future observation window period based on the historical channel state information in the observation window period and the channel state information estimation algorithm, and sort all available channels from high to low according to the channel gain; Step 5: In the first observation window of each monitoring subtask, wireless channels are allocated according to the initial priority values ​​of the called sensors, and channels with high channel gains are preferentially allocated to sensors with high initial priorities for sending monitoring data; After the first observation window period ends, the wireless channels are allocated according to the dynamic priorities obtained in step 2 and step 3, and channels with good channel status are preferentially allocated to sensors with high dynamic priorities for sending monitoring data; Step 6: The sensor sends data in multiple time slots within a transmission cycle according to the allocated wireless channel, and determines whether the data is successfully transmitted based on the ACK signal to decide whether to retransmit; before each monitoring subtask is completed, return to step 2 until the next monitoring subtask starts.

2. The method for dynamic collaborative management of industrial wireless networks for serial monitoring tasks according to claim 1, characterized in that: In step 1, for all monitoring subtasks of the overall monitoring task, the number of times each sensor is called in all monitoring subtasks is counted, and based on the statistical values, the initial priority values ​​of the sensors are set and normalized from high to low, and the priorities of the sensors are divided based on the time scale of the overall monitoring task. In the subsequent network resource allocation process, reference indicators are provided for the priorities of sensors that are called more times, so as to achieve coordination among multiple monitoring subtasks.

3. The method for dynamic collaborative management of industrial wireless networks for serial monitoring tasks according to claim 1, characterized in that: In step 2, during any monitoring subtask, an observation window is set, and the edge server calculates the average monitoring error of each sensor through the packet loss situation of the sensor and the monitoring system model; by setting the observation window, the cost of the communication resource allocation adjustment process is reduced while improving the dynamic resource allocation capability.

4. The method for dynamic collaborative management of industrial wireless networks for serial monitoring tasks according to claim 1, characterized in that: In step 2, according to the actual transmission situation of each sensor, the edge server counts the packet loss rate of each sensor, and calculates the monitoring error caused by a single transmission according to the measurement error of each sensor and the system model, which is recorded as the average monitoring error.

5. The method for dynamic collaborative management of industrial wireless networks for serial monitoring tasks according to claim 1, characterized in that: In step 2, the higher the packet loss rate of the sensor and the higher the sensor measurement error, the greater the average monitoring error.

6. The method for dynamic collaborative management of industrial wireless networks for serial monitoring tasks according to claim 1, characterized in that: In step 3, the sensor dynamic priority index calculation method sets weights to perform weighted fusion of the initial priority value of each sensor and the average monitoring error, and the weights of the initial priority value and the average monitoring error are both positive numbers and their sum is 1; the index based on the number of times the sensor is called is considered from the time scale of the total monitoring task, and the average monitoring error index due to transmission packet loss and sensor measurement error is considered on the time scale of a monitoring subtask, and the two indicators are normalized and weightedly fused.

7. The method for dynamic collaborative management of industrial wireless networks for serial monitoring tasks according to claim 1, characterized in that: In step 3, different fusion weights reflect the importance between the overall monitoring task and the individual monitoring subtasks.

8. The method for dynamic collaborative management of industrial wireless networks for serial monitoring tasks according to claim 1, characterized in that: In step 4, the channel state information estimation algorithm estimates the available channel state of a future observation window based on the historical channel state information after the first observation window of any monitoring subtask.

9. The method for dynamic collaborative management of industrial wireless networks for serial monitoring tasks according to claim 1, characterized in that: In step 5, based on the dynamic priority indicator, wireless channels are allocated. During the first observation window of each monitoring subtask, since the average monitoring error of all sensors is 0, the edge server allocates wireless channels according to the initial priority value of the sensor in the monitoring subtask, and allocates the best channel in the available wireless channels to the sensor with the highest initial priority value, thereby ensuring that sensors that are called more times will be preferentially allocated wireless channels with higher channel gains.

10. The method for dynamic collaborative management of industrial wireless networks for serial monitoring tasks according to claim 1, characterized in that: In step 5, based on the dynamic priority indicator, wireless channels are allocated. At the beginning of the second observation window of each monitoring subtask, the edge server allocates wireless channels according to the dynamic priority values ​​of the sensors in the task, and allocates the best channel in the available wireless channels to the sensor with the highest dynamic priority, thereby improving the transmission success rate and reducing the number of retransmissions and transmission energy consumption.

Images