Method for optimizing quality of service of a wireless sensor network for monitoring tension of a flight chain

Abstract

The application relates to a service quality optimization method of a scraper chain tension monitoring wireless sensor network, aiming to guarantee the energy and data queue stability of the monitoring network and meet the high-quality service requirement of the scraper chain tension monitoring network. The method combines the operation characteristics of the scraper chain, uses an improved Lyapunov drift plus punishment technology, respectively designs a punishment factor, a grade weight factor and a delay factor, guarantees the stability of the energy and data queue of the monitoring network, and realizes the continuous monitoring of the scraper chain tension.

Description

Technical Field

[0001] This invention relates to the fields of scraper chain tension monitoring and wireless sensor networks, specifically designing a quality of service optimization method for a scraper chain tension monitoring wireless sensor network based on hierarchical queue perception. Background Technology

[0002] Intelligent coal mining has become an inevitable choice for driving the energy revolution. Scraper conveyors are currently the only transportation equipment in longwall coal mining faces. Due to their prolonged operation under heavy loads in harsh environments, they are prone to chain jamming and breakage. Statistics show that scraper chain failures account for approximately 40% of scraper conveyor malfunctions. These failures can cause economic losses or even serious injuries and fatalities. Therefore, research on scraper conveyor chain condition monitoring methods is of great significance for ensuring coal mine transportation safety and improving coal mining efficiency.

[0003] Existing methods for monitoring the operating status of scraper conveyors primarily involve deploying scraper chain tension monitoring devices to monitor the chain tension. However, due to the mobility and harsh operating conditions of scraper conveyors, traditional wired monitoring methods are difficult to apply. In recent years, wireless sensor networks have played a vital role in various industrial applications due to their advantages such as flexible deployment and self-powering. They enable online monitoring of the status of key equipment and have become an important technical means for health monitoring of coal mine electromechanical equipment groups.

[0004] Currently, quality of service (QoS) optimization methods for energy-harvesting wireless sensor networks (WSNs) mainly focus on two aspects: one is for WSNs with low energy harvesting levels, which focuses more on network energy consumption; the other is for WSNs with high energy harvesting levels, which focuses more on network QoS. Among these, Lyapunov optimization-based methods have become a research hotspot due to their low computational complexity, lack of need for prior knowledge of the system's probabilistic and statistical characteristics, and ability to solve optimization problems with long-term objective functions or continuous variable constraints. However, current optimization strategies are difficult to directly apply to scraper chain tension monitoring networks, mainly because: 1) the tension monitoring data levels differ at different operating positions of the scraper chain; 2) the maximum delay requirements differ for different levels of monitoring data; and 3) the priority and maximum delay requirements of different data levels change with the periodic changes in the scraper chain's operating position.

[0005] In summary, while existing work helps optimize sensor network energy stability, data priority, and maximum latency, it still falls short of the operating conditions of scraper conveyors, thus failing to simultaneously meet the energy and quality-of-service (QoS) requirements of scraper conveyor tension monitoring networks. Therefore, this invention proposes a QoS optimization method for scraper conveyor tension monitoring wireless sensor networks based on hierarchical queue sensing. This method, while ensuring network energy stability, classifies data priorities according to different levels, meets the maximum latency and minimum packet loss requirements for different levels of monitoring data, and ultimately achieves high-quality, continuous monitoring of the scraper conveyor's operating status. Summary of the Invention

[0006] This invention proposes a quality of service optimization method for a wireless sensor network for scraper chain tension monitoring based on hierarchical queue perception. This method aims to ensure high-quality service of the scraper chain tension monitoring network while ensuring the stability of the monitoring network's energy and data queues.

[0007] A method for optimizing the quality of service (QoS) of a wireless sensor network for scraper chain tension monitoring based on hierarchical queue perception includes the following steps:

[0008] (I) Establishing a wireless sensor network system model for scraper chain tension monitoring based on built-in self-powered stress sensor nodes, specifically including:

[0009] Each self-powered stress sensor node is equipped with a piezoelectric vibration energy harvesting device and a strain gauge. The piezoelectric vibration energy harvesting device provides energy to the self-powered stress sensor node; the strain gauge is used to acquire tension data of the scraper chain. Grooves are provided on both sides of the scraper, and the self-powered stress sensor node is embedded in the grooves. A wiring groove is provided at the upper end of the scraper, and the strain gauge is fixed to the link of the scraper chain corresponding to the scraper. The wire of the strain gauge is embedded in the wiring groove and connected to the data acquisition module of the self-powered stress sensor node to acquire the tension data of the scraper chain in real time. Data sampled in the upward phase of the scraper is defined as high-level data, and data sampled in the downward phase of the scraper is defined as low-level data. The transmission priority of the monitoring data in the upward and downward phases of the scraper should meet the following conditions:

[0010] ,

[0011] in, Indicates the priority of higher-level data; These represent the priority of lower-level data; This means that when the monitoring data exceeds the threshold or an anomaly occurs, the priority of the monitoring data collected by this node is set to the highest level to ensure that the abnormal data is sent first.

[0012] Establish a scraper chain tension monitoring network model and define For the aggregation node set, The monitoring node set consists of 4 monitoring nodes and 2 aggregation nodes for data transmission. The position of the aggregation nodes is fixed, while the position of the monitoring nodes changes periodically with the movement of the scraper. The monitoring nodes transmit tension data to the aggregation nodes in a single-hop manner, and the aggregation nodes can only receive data from one monitoring node at a time. Each node has two data queues, which are used to store high-level data collected by the monitoring nodes during the upward phase of the scraper chain and low-level data collected during the downward phase.

[0013] (II) Based on the wireless sensor network system model for scraper chain tension monitoring established in step (I), establish a monitoring network data transmission model, specifically including:

[0014] In step (1), power is allocated to each monitoring node. Used for data transmission; time slot The transmission power allocation matrix is ,in, It is a link The power allocated above, Constraints must be met:

[0015] ,

[0016] Among them, L n To monitor the links corresponding to the nodes, where L is the total number of links in the network, and P... max Maximum transmission power;

[0017] Define the link state set as ,in It is a time slot The channel states of the time link are independently distributed, and the link capacity is... At that time, the link There is an upper bound between the data capacity and the allocated transmission power. The relationship between the positive constants, where the positive constants are... Given the limitation on node transmission rate under unit power conditions, based on the uplink and downlink data classification method in step (i), the optimized node transmission rate is expressed as:

[0018] ,

[0019] In the formula, This indicates the actual transmission rate of high-level data; This indicates the actual transmission rate of low-level data.

[0020] When higher-priority data has a higher priority, it is sent first; otherwise, lower-priority data is sent. When the transmission rate... Greater than When the queue length for the graded data is reached, the remaining channel capacity will be used to send data for other grades.

[0021] (III) Based on the monitoring network data transmission model established in step (II), establish a monitoring network energy queue model, specifically including:

[0022] The dynamic update process of the energy queue backlog length of the self-powered stress sensor node is represented as:

[0023] ,

[0024] in, This represents the duty cycle of the node at the current time, and , As a disturbance factor, In order to collect power, This refers to the energy queue backlog length.

[0025] Transmission power To satisfy the surplus energy demand constraint, namely:

[0026] ;

[0027] Minimum Remaining Energy , is represented as:

[0028] ;

[0029] (iv) Based on the monitoring network energy queue model established in step (iii), establish data queue models at different levels, specifically including:

[0030] definition For time slots Arrived inside The number of data packets for the grade data. To monitor nodes in time slots Waiting to send Queue length and transmission rate of graded data It's about transmission power. and channel state Continuous functions, Indicates monitoring node In the link The data transmission rate above, then the first The data queue length of each monitoring node at the next moment is represented as follows:

[0031] ,

[0032] Wherein, the initial value of the queue length =0, and These represent the high-level and low-level data queues in the time slots, respectively. The number of packet losses, and the following:

[0033] ,

[0034] in, Discarded due to exceeding maximum latency requirements The maximum number of packets lost in the graded data queue;

[0035] The data queue backlog length must meet the following constraints:

[0036] ;

[0037] A time-delay queue model was introduced to establish a time-delay queue model suitable for scraper chain condition monitoring networks. The process of updating the queue length can be represented as:

[0038] ,

[0039] in, It is a pre-defined constant, and ;

[0040] Delay queue Stability requirements must be met:

[0041] ;

[0042] The maximum latency of data in different queue levels is expressed as... ,and:

[0043] ,

[0044] in, express The actual maximum delay of the grade monitoring data; express Theoretical maximum delay of graded monitoring data;

[0045] (v) Based on the data queue models of different levels established in step (iv), establish a model for optimizing packet loss based on the level queues, specifically including:

[0046] Each time slot Taking into account energy stability constraints, the equation is defined as follows: The priority constraint formula for the transmission of data at different levels is defined as follows: The maximum delay constraint formula for different levels of data is defined as follows: The transmission power constraint formula is defined as follows: The required transmission rate for different levels of data is defined as follows: The network stability requirement is defined as follows: , , and The mathematical description of the optimization problem is as follows:

[0047] ,

[0048] In the formula, express The delay factor for graded data can be used to control the maximum delay of different grades of data to meet the different delay requirements of high and low grade data.

[0049] (vi) Based on the packet loss quantity optimization problem model established in step (v) based on the hierarchical queue, a penalty factor is introduced. Rank weight factor and delay factor An improved Lyapunov drift penalty function is constructed, specifically including:

[0050] Introducing grade weighting factors Furthermore, when monitoring data anomalies occur, the monitoring data level weight factor for that node is adjusted. It will be set to the highest priority When the ranking weight factor is multiplied by the actual queue or delayed queue, the queue model... Adjusted to ,make for , and The set vector, assuming time slots The improved Lyapunov function is obtained:

[0051] ,

[0052] This includes information on the queue lengths of all monitoring nodes in the network, resulting in the improved Lyapunov drift as follows:

[0053] ,

[0054] By each time slot Greedily approaching the Lyapunov drift to a value close to zero, i.e., minimizing the Lyapunov drift. This can meet the network stability requirements;

[0055] Based on the principle of drift penalty, the problem of optimizing the number of lost packets is transformed into the problem of minimizing the upper limit of the penalty function for each time slot, and a non-negative control parameter is introduced. The improved Lyapunov drift penalty function is expressed as:

[0056] ,

[0057] in, , and These represent the penalty factor, delay factor, and grade weight factor, respectively.

[0058] when When set to zero, the improved Lyapunov drift penalty function is equivalent to the problem of minimizing Lyapunov drift; when When the values ​​are normal, the system balances queue backlog, data priority, and packet loss in the monitoring network while ensuring the stability of the energy queue and data queue.

[0059] (vii) Based on the improved Lyapunov drift penalty function constructed in step (vi), a network service quality optimization algorithm based on hierarchical queue awareness is proposed. Distributed algorithms are used to optimize the sub-problems of energy management, data reception, delay guarantee, power allocation, data transmission, and queue update, so as to meet the service quality requirements of the scraper chain tension monitoring network.

[0060] Preferably, in step (i), the strain gauge is fixed to the middle of the link of the scraper chain, and the link is a circular link.

[0061] Preferably, in step (iii), a virtual queue is introduced to establish a monitoring network energy queue model. Data exceeding the maximum latency of a node is dropped to meet the maximum latency requirement of the monitoring data. The virtual queue length of each monitoring node at the next moment is represented as:

[0062] ,

[0063] in, This is an indicator function; it takes a value of 1 when the monitoring node moves to a different level of region, and 0 otherwise. Furthermore, the virtual queue needs to meet stability requirements:

[0064] .

[0065] Preferably, the problem of minimizing the long-term packet loss under constraints is transformed into a problem of minimizing drift penalty in a single time slot, in any time slot. Given feasible solutions for energy management, delay guarantees, power allocation, and data transmission, the improved Lyapunov drift penalty function in step (vi) has the following upper bound:

[0066] ,

[0067] In the formula, It is a finite positive constant that satisfies the following condition:

[0068] .

[0069] Preferably, in step (vii), energy management is optimized by adjusting the duty cycle to achieve node energy balance; data reception is optimized by determining the data reception rate according to the operating stage of the scraper chain; delay guarantee is optimized by handling packet loss for data that is about to exceed the maximum delay; power allocation is optimized by allocating transmission power according to the level weight factor and queue length; data transmission is optimized by prioritizing the transmission of level data with higher importance factors; and queue updating is optimized by updating the queue length according to the queuing dynamics formula.

[0070] Preferably, when optimizing energy management, the energy state of the node is considered. By adjusting the duty cycle This is to achieve energy balance at the nodes; when a node's energy is insufficient, reduce... Otherwise, increase To achieve energy balance at the nodes, the change in duty cycle at the next time step is expressed as:

[0071] ;

[0072] Nodes in their remaining energy Exceeding the minimum energy required for its operation It will run when needed; otherwise, it will remain in hibernation, waiting for energy replenishment. A fault indicator factor has been introduced. That is, when the tension monitoring data shows abnormalities, the fault indicator factor defaults to 0. At this time, the node's Automatically set to 1 to ensure timely transmission of abnormal data. Represented as:

[0073] .

[0074] Preferably, when optimizing data reception, the monitoring node is in the time slot Maximum data reception rate at time Based on the current time slot duty cycle Maximum sampling frequency The receiving rate of high-level data is determined jointly by the scraper chain operation phase; when the monitoring node is in the scraper uplink phase, the receiving rate is expressed as... Low-level receiving rate is expressed as Conversely, when the monitoring point is in the downward phase of the scraper, the reception rate of high-level data is expressed as... Low-level receiving rate is expressed as .

[0075] Preferably, when optimizing the delay guarantee, in the time slot Within this process, data that is about to exceed the maximum delay is packet-dropped to ensure the validity of the monitoring data;

[0076] The optimization problem can be expressed as:

[0077] ;

[0078] The optimal solution is:

[0079] ,

[0080] in, express The tiered data queue in the time slot The number of packets lost on the network As a penalty factor, As the delay factor, This is the length of the virtual queue. For delayed queues, Discarded due to exceeding maximum latency requirements The maximum number of packets lost in the graded data queue. express The actual maximum delay of the grade monitoring data, express The theoretical maximum latency of the level monitoring data; when the amount of data in the queue exceeds a set threshold, the timeliness of the monitoring data will be ensured by discarding data at the tail of the queue, in order to prevent the monitoring data from becoming invalid due to exceeding the maximum latency, by adjusting the penalty factor. Change the maximum latency for all levels of data by adjusting the latency factor. Achieve maximum delay control for high- and low-level data.

[0081] Preferably, when optimizing power allocation, in any time slot In the scraper chain tension monitoring network, there are two or fewer monitoring nodes that send data to two aggregation nodes respectively; within each time slot, each aggregation node can only receive data from one monitoring node in its area, meaning there is only one monitoring node. Allocated transmission power, the rest Each node is in sleep mode, taking into account the level weighting factor. Transmission rate and queue length The node's transmission power allocation strategy is expressed as:

[0082] ,

[0083] in, Indicates the aggregation node The number of monitoring nodes in the area; when monitoring data anomalies occur, the priority factor of the monitoring data collected by that node will be set to the highest priority. Meanwhile, the transmission power is set to [default value]. This ensures timely early warning of scraper chain malfunctions.

[0084] Preferably, when optimizing data transmission, the actual latency level is utilized. and rank weighting factor Determine the importance of the data; assume the importance factor is... ,in This serves as a data flow control factor, determined by comparing factors of varying importance. The size is used to determine the current time slot. The data type that should be sent should be prioritized, with data of higher importance factors being transmitted first; when the transmission rate... Greater than the currently transmitted Queue backlog length of graded data At that time, in the remaining channel capacity Transmit other levels of data within the channel to improve channel utilization.

[0085] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0086] 1. This invention combines the operating characteristics of scraper chains and proposes a quality of service optimization method for wireless sensor networks for scraper chain tension monitoring based on Lyapunov optimization drift plus penalty technology. This method ensures the stability of the monitoring network energy and data queue, thereby achieving high-quality and sustainable monitoring of scraper chain tension.

[0087] 2. The penalty factor designed in this invention This allows the system to balance the backlog and packet loss of data queues at different levels of the monitoring network while ensuring the stability of the energy queue and data queue, thereby minimizing the number of packet losses in scraper chain tension monitoring data.

[0088] 3. This invention introduces a rank weight factor into the traditional Lyapunov optimization framework. This parameter enables priority scheduling of data at different levels to meet the priority requirements of different data levels.

[0089] 4. To meet the delay requirements of monitoring data at different levels, this invention designs a level delay factor. It can adjust the maximum delay for different levels of data.

[0090] 5. The network service quality optimization method based on hierarchical queue awareness proposed in this invention can be extended to other rotating machinery condition monitoring networks with different data priorities and latency requirements. Attached Figure Description

[0091] Figure 1 This is a flowchart of the method of the present invention;

[0092] Figure 2 This is the scraper chain tension monitoring network model of the present invention;

[0093] Figure 3 This is a flowchart of the queue update process of the present invention;

[0094] Figure 4 This invention relates the average number of packet losses in the network to the penalty factor.

[0095] Figure 5 This invention relates the average queue length of network data to the penalty factor.

[0096] Figure 6 This invention relates the maximum network data latency to the penalty factor.

[0097] Figure 7 This refers to the situation where the queue backlog length of the present invention changes with the time slot;

[0098] Figure 8 This invention relates the average number of packet losses in network data to the grade weighting factor.

[0099] Figure 9 This invention relates the average queue backlog length of network data to the grade weighting factor.

[0100] Figure 10 This invention relates the maximum network data latency to the grade weighting factor.

[0101] Figure 11 This invention relates the average number of packet losses in the network to the latency factor.

[0102] Figure 12 This invention relates the average queue length of network data to the delay factor.

[0103] Figure 13 This relates the maximum network data latency to the latency factor in this invention. Specific implementation methods

[0104] To make the objectives, technical solutions, and optimizations of this embodiment clearer and more explicit, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0105] like Figure 1As shown, this embodiment provides a method for optimizing the quality of service (QoS) of a wireless sensor network for scraper chain tension monitoring based on hierarchical queue perception, specifically including:

[0106] Step 1: Establish a wireless sensor network system model for scraper chain tension monitoring based on built-in self-powered stress sensing nodes;

[0107] like Figure 2 As shown, the specific steps are as follows:

[0108] Each self-powered stress sensing node is equipped with an energy harvesting device and a strain gauge. The energy harvesting device provides energy to the stress sensor node, and the strain gauge acquires the tension data of the scraper chain. The deployment of the stress sensor node is as follows: First, grooves are made on both sides of the scraper, and the monitoring node is embedded in the grooves on both sides of the scraper; then, a wiring groove is designed at the upper end of the scraper, and the strain gauge is fixed to the optimal monitoring position of the circular chain stress of the scraper chain, i.e., the middle of the chain link; finally, the wire of the strain gauge is embedded in the wiring groove and connected to the data acquisition module of the stress sensor node to acquire the tension data of the scraper chain in real time.

[0109] The data acquisition module in this embodiment is a part of the existing module known to those skilled in the art, which is a stress sensor node.

[0110] The scraper groove and wiring groove described in this embodiment are made using existing methods well known to those skilled in the art, and will not be described in detail here.

[0111] The energy harvesting device and strain gauge in this embodiment are existing products or structures well known to those skilled in the art, and will not be described in detail here.

[0112] Because the scraper chain not only carries the coal to be transported during the upward phase but is also susceptible to impact from gangue and other debris, data sampled during the upward phase is more useful for fault detection than data sampled during the downward phase. Therefore, data sampled during the upward phase is defined as high-level data, and data sampled during the downward phase is defined as low-level data. The transmission priority of monitoring data during the upward and downward phases should then meet the following conditions:

[0113] ,

[0114] in, Indicates the priority of higher-level data; These represent the priority of lower-level data; This means that when the monitoring data exceeds the threshold or an anomaly occurs, the priority of the monitoring data collected by this node will be set to the highest level to ensure that the abnormal data is sent first.

[0115] Based on the above, a scraper chain tension monitoring network model is established. Since the convergence nodes are usually located at both ends of the hydraulic support, a [model name] is defined. For the aggregation node set, This is a set of monitoring nodes. Four monitoring nodes and two aggregation nodes are configured for data transmission. The positions of the aggregation nodes are fixed, while the positions of the monitoring nodes change periodically with the movement of the scraper. The monitoring nodes transmit tension data to the aggregation nodes in a single-hop manner, and each aggregation node can only receive data from one monitoring node at a time. Furthermore, this embodiment assumes that each node has two data queues, one for storing high-level data collected during the upward phase of the scraper chain and the other for storing low-level data collected during the downward phase.

[0116] Step 2: Based on Step 1, establish a monitoring network data transmission model;

[0117] The specific steps are as follows:

[0118] In order to transmit monitoring data to the destination, each monitoring node needs to be allocated power. Used for data transmission; assuming a time slot The transmission power allocation matrix is ,in, It is a link The power allocated above, Constraints must be met: ;

[0119] Define the link state set as ,in, It is a time slot The channel states of the link are independent and distributed. Assume the link capacity is... Then the link There is an upper bound between the data capacity and the allocated transmission power. The relationship between the positive numbers. Given the limitation on node transmission rate under unit power conditions, and based on the uplink / downlink data tier classification method, the optimized node transmission rate is expressed as:

[0120] ,

[0121] In the formula, and These represent the actual transmission rates of high-level and low-level data, respectively.

[0122] When higher-priority data has a higher priority, it is sent first; otherwise, lower-priority data is sent. When the transmission rate... Greater than When the queue length for the graded data is reached, the remaining channel capacity will be used to send data for other grades.

[0123] Step 3: Based on Step 2, establish a monitoring network energy queue model;

[0124] The specific steps are as follows:

[0125] The dynamic update process of the energy queue backlog length of a node is represented as follows:

[0126] ,

[0127] in, This represents the duty cycle of the node at the current time, and , For disturbance factor;

[0128] Considering that the energy consumed by a node in any time slot must not exceed the node's current remaining energy, i.e., the transmission power. If the surplus energy demand constraint must be satisfied, then:

[0129] ;

[0130] To ensure the continuity of the monitoring network, the energy queue backlog length of a node should not be less than the minimum remaining energy required for the node to operate. , is represented as:

[0131] .

[0132] Step 4: Based on Step 3, establish data queue models at different levels;

[0133] The specific steps are as follows:

[0134] Transmission rate It's about transmission power. and channel state Continuous functions, using To represent monitoring nodes In the link The data transmission rate is defined. For time slots Arrived inside The number of data packets for the grade data. To monitor nodes in time slots Waiting to send The queue length of the graded data, then the... The data queue length of each monitoring node at the next moment is represented as follows:

[0135] ,

[0136] Wherein, the initial value of the queue length =0, and These represent the high and low level data queues in the time slots, respectively. The number of packet losses, and the following:

[0137] ,

[0138] in, Discarded due to exceeding maximum latency requirements The maximum number of packets lost in the graded data queue;

[0139] To ensure the stability of the monitoring network, the data queue backlog length must meet the following constraints:

[0140] ;

[0141] The periodic movement of the scraper chain tension monitoring nodes causes periodic changes in the level of monitoring data. Specifically, when a moving monitoring node moves from the uplink area to the downlink area, the arrival rate of high-level data becomes 0. Similarly, when a moving monitoring node moves from the downlink area to the uplink area, the arrival rate of low-level data becomes 0. Therefore, a virtual queue is introduced to drop packets of data that have exceeded the node's maximum delay, thus meeting the maximum delay requirement for monitoring data. Then the... The virtual queue length of each monitoring node at the next moment is further represented as:

[0142] ,

[0143] in, This is an indicator function; it takes a value of 1 when the monitoring node moves to a different level of region, and 0 otherwise. Furthermore, the virtual queue needs to meet stability requirements:

[0144] ;

[0145] Building upon the above, to ensure that the delay of monitoring data is bounded under worst-case conditions, a delay queue model is introduced, and the time-averaged delay constraint is further transformed into a queue stability problem. This embodiment establishes a delay queue model adapted to the scraper chain operating condition monitoring network. The process of updating the queue length is represented as follows:

[0146] ,

[0147] in, It is a pre-defined constant, and ;

[0148] To ensure the stability of the monitoring network, a delay queue is used. Stability requirements must be met:

[0149] ;

[0150] When the actual queue backlog is not zero, it indicates that there is still data in the queue that has not been served for a long time. This will increase the maximum latency of different levels of queue data proposed in this embodiment, which is expressed as follows: Considering that the timeliness of monitoring data directly affects the safety of scraper chain operation, and that the more timely the monitoring data is sent, the higher the accuracy of fault condition early warning, and that high-level data should have lower latency, this is expressed as follows:

[0151] ,

[0152] in, express The actual maximum delay of the grade monitoring data; express Theoretical maximum delay of graded monitoring data;

[0153] This assumes the existence of an algorithm that can control the stability of the monitoring network system and ensure... and ,in, and Both are bounded. If the service and discard processes follow the FIFO principle, then all items in the queue... Maximum latency of graded data Represented as:

[0154] ;

[0155] The proof of the above theorem is as follows:

[0156] For any time slot The earliest time slot for the monitoring data to arrive is the earliest time slot to leave the queue. At the same time, all time slots Data arriving within the time slot It is either served previously or discarded. If this assumption is not true, the following contradiction arises: for all time slots... They all Otherwise, the queue data will be cleared. In the time slot It contains:

[0157] ;

[0158] Based on the above formula, for time slots Summing the queues within the queue, we get:

[0159] ;

[0160] Rearranging the above formula, since... , Then we have:

[0161] ;

[0162] According to queuing dynamics theory, in time slots Arrived data It will be placed at the end of the queue, and only if the queue... The data will only be served after all the backlog has been cleared. That is, the last arriving data... In the time slot Leave, among them And to satisfy The smallest integer of the function. Assume... exist If you haven't been served before, then... Substituting it into the above formula, we get:

[0163] ;

[0164] After reorganization, it can be seen that this contradicts the assumption given in the above theorem, so the theorem is proved.

[0165] Step 5: Based on Step 4, establish a model for optimizing packet loss based on hierarchical queues;

[0166] The specific steps are as follows:

[0167] Each time slot Taking into account energy stability constraints Priority constraints for data transmission at different levels Maximum delay constraints for different levels of data Transmission power constraint type Different levels of data transmission rate requirements Network stability requirements , , and Therefore, the mathematical description of the optimization problem can be obtained as follows:

[0168] ,

[0169] In the formula, express Delay factors for graded data. These delay factors allow for the control of the maximum delay for different grades of data, thus meeting the varying delay requirements of high- and low-grade data.

[0170] Step Six: Based on Step Five, introduce a penalty factor. Rank weight factor and delay factor Construct an improved Lyapunov drift penalty function;

[0171] The specific steps are as follows:

[0172] Based on the traditional Lyapunov function, a rank weighting factor is introduced. Furthermore, when monitoring data anomalies occur, the monitoring data level weight factor for that node is adjusted. It will be set to the highest priority If the rank weight factor is directly multiplied by the actual queue or delayed queue, then the queue model will have... This item indicates that when the rank weight factor Increasing the packet loss rate leads to a simultaneous increase in both transmission rate and packet loss, which is clearly unreasonable. For high-priority data, packet loss should be minimized to achieve a higher transmission rate. Conversely, the transmission rate of low-level data... It should be lower. Therefore, the above formula should be adjusted to: That is, the grade weighting factor Transmission rates for high and low priority data only Adjustments can be made without affecting the number of packet losses. .

[0173] Based on this, let for , and The set vector, assuming time slots Then we obtain the improved Lyapunov function:

[0174] ,

[0175] This includes queue length information for all monitoring nodes in the network. Further, an improved Lyapunov drift can be derived as follows:

[0176] ,

[0177] By each time slot Greedily approaching the Lyapunov drift to a value close to zero, i.e., minimizing the Lyapunov drift. This can meet the network stability requirements;

[0178] Based on the drift penalty principle, the problem of optimizing the number of lost packets can be transformed into minimizing the upper bound of the penalty function for each time slot. A non-negative control parameter is introduced. The improved Lyapunov drift penalty function is then expressed as:

[0179] ,

[0180] in, , and These represent the penalty factor, delay factor, and rank weight factor, respectively, which are constantly changing. , and In reality, it's constantly adjusting the number of lost packets and the queue length, so it can be... , and It can be considered as a weighting factor for the number of lost packets and the level of the data queue;

[0181] Among them, by adjusting It can also achieve differentiated control over the number of packet losses for high- and low-level data to meet the different latency requirements of different data levels. When the value is set to zero, the above equation is equivalent to minimizing Lyapunov drift; while when When the values ​​are normal, the system balances queue backlog, data priority, and packet loss in the monitoring network while ensuring the stability of the energy queue and data queue.

[0182] This embodiment transforms the problem of minimizing the long-term packet loss under constraints into a problem of minimizing drift penalty in a single time slot, in any time slot. Given feasible solutions for energy management, delay guarantees, power allocation, and data transmission, the Lyapunov drift penalty function has the following upper bound:

[0183] ,

[0184] In the formula, It is a finite positive constant that satisfies the following condition:

[0185] ;

[0186] The proof is as follows:

[0187] Based on the inequality property, for any , , For all of them, the following inequalities hold:

[0188] ,

[0189] According to the above formula, we have:

[0190] 1) For virtual queues :

[0191] because , , Then we have:

[0192] ;

[0193] Similarly, for virtual queues ,have:

[0194] ;

[0195] 2) For delayed queues :

[0196] because , , Then we have:

[0197] ;

[0198] Similarly, for delayed queues ,have:

[0199] ;

[0200] 3) For energy queues :

[0201] because , , Then we have:

[0202] ;

[0203] The proof can be obtained by rearranging the above inequalities.

[0204] Step 7: Based on Step 6, a network service quality optimization algorithm based on hierarchical queue awareness is proposed. A distributed algorithm is used to optimize the sub-problems of energy management, data reception, delay guarantee, power allocation, data transmission, and queue update, so as to meet the service quality requirements of the scraper chain tension monitoring network.

[0205] The specific steps are as follows:

[0206] Energy management is based on the energy state of nodes. By adjusting the duty cycle To achieve energy balance at the nodes;

[0207] To ensure the stability of the monitoring node's energy, this embodiment proposes an adaptive node duty cycle strategy, which is based on the node's energy state. By adjusting the duty cycle This is to achieve energy balance at the nodes. When a node's energy is insufficient, the energy is reduced. Otherwise, increase This is to achieve energy balance at the nodes. The change in duty cycle at the next moment is expressed as:

[0208] ;

[0209] A node can only function if it has remaining energy. Exceeding the minimum energy required for its operation It can only operate when the power is available; otherwise, it will remain in a dormant state, waiting for energy replenishment. Furthermore, to ensure timely fault warnings, this invention introduces a fault indicator factor. That is, when the tension monitoring data shows abnormalities, the fault indicator factor defaults to 0. At this time, the node's Automatically set to 1 to ensure timely transmission of abnormal data. Represented as:

[0210] ;

[0211] Monitoring nodes in time slots Maximum data reception rate at time Based on the current time slot duty cycle Maximum sampling frequency The operation phase of the scraper chain is determined together;

[0212] When the monitoring node is in the uplink phase, the reception rate of high-level data is expressed as: Low-level receiving rate is expressed as Conversely, when in the downlink phase, the reception rate of high-level data is expressed as... Low-level receiving rate is expressed as ;

[0213] In the time slot Within this process, data that is about to exceed the maximum delay will be dropped to ensure the validity of the monitoring data.

[0214] The optimization problem can be expressed as:

[0215] ;

[0216] It can be seen that this is a linear programming problem, and the optimal solution is:

[0217] ,

[0218] in, As the delay factor, This is the length of the virtual queue. This is a delayed queue; when the amount of data in the queue exceeds a certain set threshold (the threshold in this embodiment is one that can be set by those skilled in the art based on actual conditions), the timeliness of the monitoring data will be ensured by discarding data at the tail of the queue, thus preventing the monitoring data from becoming invalid due to exceeding the maximum delay. Therefore, the penalty factor is adjusted... The maximum latency for all data levels can be changed by adjusting the latency factor. The maximum delay of high and low level data can be controlled according to the actual monitoring service needs; the maximum delay in this embodiment is the maximum delay that can be determined or set by those skilled in the art based on the actual situation.

[0219] In any time slot In the scraper chain tension monitoring network, a maximum of two monitoring nodes send data to two aggregation nodes respectively; within each time slot, each aggregation node can only receive data from one monitoring node within its area, meaning there is only one monitoring node. Allocated transmission power, the rest Each node is in sleep mode. The weighting factor is taken into account. Transmission rate and queue length The node's transmission power allocation strategy is expressed as:

[0220] ,

[0221] in, Indicates the aggregation node The number of monitoring nodes in the area; when monitoring data anomalies occur, the priority factor of the monitoring data collected by that node will be set to the highest priority. Meanwhile, the transmission power is set to [default value]. This is to ensure timely early warning of scraper chain malfunctions;

[0222] To ensure high priority for high-level data transmission, the actual latency is utilized. and rank weighting factor Determine the importance of the data;

[0223] Assuming the importance factor is ,in This serves as a data flow control factor, determined by comparing factors of varying importance. The size is used to determine the current time slot. The data types that should be sent should be prioritized, with data of higher importance factors being transmitted first.

[0224] When the transmission rate Greater than the currently transmitted Queue backlog length of graded data At that time, in the remaining channel capacity Transmit other levels of data within the channel to improve channel utilization.

[0225] In this application embodiment, the present invention provides a QoS optimization method for a monitoring network based on graded queue awareness, the process of which is as follows: Figure 3 As shown, it includes the following steps:

[0226] First, initialize the parameters. , and ,initialization When it is 1, Less than or equal to At that time, the following steps are repeated:

[0227] based on Obtain the optimal duty cycle of the node Calculate the actual delay of the current time slot monitoring node queue data. According to the formula Data that is about to time out is dropped, and the monitoring nodes estimate the channel state respectively. And obtain the transmission rate at this time. According to the formula Obtain the transmission power of all monitoring nodes in the current area. ;

[0228] if According to And combined with importance factors Obtain the actual transmission rate of the monitoring node It also sends scraper chain tension monitoring data to the aggregation nodes within its region;

[0229] Otherwise monitoring node Enter sleep mode;

[0230] According to the formula , , , Update the queues respectively , , and .

[0231] The proposed network service quality optimization algorithm based on hierarchical queue awareness in this embodiment studies different penalty factors through theoretical numerical simulation. Rank weight factor and delay factor Impact on the performance of the scraper chain tension monitoring network:

[0232] This embodiment first studies different penalty factors. The specific impact on network performance is as follows:

[0233] 1) This embodiment studies different penalty factors. Monitor the changes in the average number of packet losses per node in real time. For example... Figure 4 As shown, with the penalty factor As the penalty factor continues to increase, the average number of packet losses across the entire network first decreases rapidly, then tends to converge. The key lies in the penalty factor. The increase in the value increases the maximum latency of the monitoring data, which further reduces the chance of packet loss, thus achieving a trade-off between the data queue length and the number of packet losses.

[0234] 2) This embodiment studied different penalty factors. The change in queue backlog length during normal times, Figure 5 Different penalty factors The trend of the average queue backlog length over time shows that a larger... Values ​​that lead to greater queue congestion in the system. Furthermore, the average queue length and penalty factor... Proportional;

[0235] 3) This embodiment studies the maximum latency and penalty factor of monitoring network nodes. Relationships, such as Figure 6 As shown, under different penalty factors Under these conditions, the proposed algorithms can all meet the maximum latency requirements of the monitoring data. Furthermore, in-depth analysis reveals that with the increase in the penalty factor... As the number of packets lost increases, the actual maximum latency of the monitoring data generally shows an upward trend. This phenomenon can be attributed to the increasing number of packets lost with the increase in the penalty factor. The increase in latency leads to a decrease in latency, which in turn increases the backlog length of the data queue, thereby further increasing the maximum latency.

[0236] 4) This embodiment analyzes different time slots. Monitor the changes in the backlog length of the energy queue and data queue of the monitoring nodes in real time. Figure 7 As can be seen in (a), the backlog length of both high- and low-level data queues at the monitoring nodes has a definite upper limit, indicating that the proposed algorithm can guarantee the stability of the monitoring node data queues. Meanwhile, from Figure 7 As shown in (b), the energy queue of the monitoring node also exhibits a very stable performance, indicating that the energy balancing strategy proposed in this invention can guarantee the stability of the monitoring network energy under heavy load conditions.

[0237] This embodiment studies weighting factors of different levels. The specific impact on network performance is as follows:

[0238] 1) This embodiment applies weighting factors to different high-level data levels. The average number of packet losses at the monitored nodes was analyzed, such as... Figure 8 As shown, with the grade weight factor With the continuous increase in packet loss, the number of packet losses for high-level data during the uplink phase generally shows a decreasing trend, while the number of packet losses for low-level data increases significantly. This phenomenon indicates that adjusting the level weighting factor... The priority of monitoring data can be changed, thus ensuring the priority of high-level data. In-depth analysis shows that when the level weight factor... As the number of packets lost continues to increase, the average number of packets lost for both high- and low-level data tends to converge. This indicates that the level weight factor... There are optimal values, and in practical applications, appropriate level weighting factors can be selected according to monitoring needs;

[0239] 2) This embodiment studies the average queue backlog length and grade weighting factor of high and low grade data. Relationships, such as Figure 9 As shown, with the grade weight factor With the increase of the rank weight factor, the average queue backlog length of high-rank data generally shows a decreasing trend and slowly converges. However, the queue backlog length of low-rank data first increases and then decreases. This phenomenon is due to the rank weight factor. The increase in the number of high-level data transmission opportunities leads to a trend of increasing queue backlog length for low-level data. This is due to the increasing weighting factor of the data level. As the priority of high-level data increases further, in order to not exceed the maximum latency of low-level data, the algorithm continuously increases the number of packet losses of low-level data, thereby slowly reducing the queue backlog of low-level data and tending to converge.

[0240] 3) This embodiment studies different levels of weighting factors. The latency of monitoring data at different levels, such as Figure 10 As shown, the proposed algorithm has different levels of weight factors. Under all conditions, the maximum delay requirements for monitoring data of different levels can be met, thereby ensuring the effectiveness of the monitoring data.

[0241] This embodiment studies the impact of different delay factors on network performance, specifically:

[0242] 1) This embodiment studies different delay factors. The changes in the average number of packet losses between high and low-level data, such as Figure 11 As shown, with the low-level data delay factor As the latency increases, the number of packet losses for high-level data initially increases and then tends to converge. Conversely, the average number of packet losses for low-level data initially decreases and then tends to converge. This phenomenon indicates that a larger latency factor can further reduce the average number of packet losses at monitoring nodes.

[0243] 2) This embodiment studies the average queue backlog length of high- and low-level data as a function of the delay factor. The situation of change, such as Figure 12 As shown, the average queue length exhibits a trend of first increasing and then converging. This is mainly due to the delay factor. The increase in latency of low-level data leads to an increase in the maximum latency of low-level data, which in turn leads to a continuous increase in the average queue backlog length of low-level data. This reduces the chance of high-level data being transmitted, so the average queue backlog length of high-level data also shows an increasing trend.

[0244] 3) This embodiment analyzes different delay factors. The latency of high- and low-level data, such as Figure 13 As shown, since only the low-level data delay factor was changed... The maximum latency of high-level data remains relatively stable, while the maximum latency of low-level data generally shows an increasing trend. In-depth analysis shows that the actual maximum latency of monitoring data at different levels is less than the theoretical maximum latency, indicating that the proposed latency factor can guarantee the maximum latency requirements of data at different levels.

[0245] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications and equivalent changes made to the above embodiments based on the technical essence of the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for optimizing the quality of service (QoS) of a wireless sensor network for monitoring scraper chain tension, characterized in that, Includes the following steps: (I) Establishing a wireless sensor network system model for scraper chain tension monitoring based on built-in self-powered stress sensor nodes, specifically including: Each self-powered stress sensor node is equipped with a piezoelectric vibration energy harvesting device and a strain gauge. The piezoelectric vibration energy harvesting device provides energy to the self-powered stress sensor node; the strain gauge is used to acquire tension data of the scraper chain. Grooves are provided on both sides of the scraper, and the self-powered stress sensor node is embedded in the grooves. A wiring groove is provided at the upper end of the scraper, and the strain gauge is fixed to the link of the scraper chain corresponding to the scraper. The wire of the strain gauge is embedded in the wiring groove and connected to the data acquisition module of the self-powered stress sensor node to acquire the tension data of the scraper chain in real time. Data sampled in the upward phase of the scraper is defined as high-level data, and data sampled in the downward phase of the scraper is defined as low-level data. The transmission priority of the monitoring data in the upward and downward phases of the scraper should meet the following conditions: ， in, Indicates the priority of higher-level data; These represent the priority of lower-level data; This means that when the monitoring data exceeds the threshold or an anomaly occurs, the priority of the monitoring data collected by this node is set to the highest level to ensure that the abnormal data is sent first. Establish a scraper chain tension monitoring network model and define For the aggregation node set, The monitoring node set consists of 4 monitoring nodes and 2 aggregation nodes for data transmission. The position of the aggregation nodes is fixed, while the position of the monitoring nodes changes periodically with the movement of the scraper. The monitoring nodes transmit tension data to the aggregation nodes in a single-hop manner, and the aggregation nodes can only receive data from one monitoring node at a time. Each node has two data queues, which are used to store high-level data collected by the monitoring nodes during the upward phase of the scraper chain and low-level data collected during the downward phase. (II) Based on the wireless sensor network system model for scraper chain tension monitoring established in step (I), establish a monitoring network data transmission model, specifically including: In step (1), power is allocated to each monitoring node. Used for data transmission; time slot The transmission power allocation matrix is ,in, It is a link The power allocated above, Constraints must be met: ， Among them, L n To monitor the links corresponding to the nodes, where L is the total number of links in the network, and P... max Maximum transmission power; Define the link state set as ,in It is a time slot The channel states of the time link are independently distributed, and the link capacity is... At that time, the link There is an upper bound between the data capacity and the allocated transmission power. The relationship between the positive constants, where the positive constants are... Given the limitation on node transmission rate under unit power conditions, based on the uplink and downlink data classification method in step (i), the optimized node transmission rate is expressed as: ， In the formula, This indicates the actual transmission rate of high-level data; This indicates the actual transmission rate of low-level data. When higher-priority data has a higher priority, it is sent first; otherwise, lower-priority data is sent. When the transmission rate... Greater than When the queue length for the graded data is reached, the remaining channel capacity will be used to send data for other grades. (III) Based on the monitoring network data transmission model established in step (II), establish a monitoring network energy queue model, specifically including: The dynamic update process of the energy queue backlog length of the self-powered stress sensor node is represented as: ， in, This represents the duty cycle of the node at the current time, and , As a disturbance factor, In order to collect power, This refers to the energy queue backlog length. Transmission power To satisfy the surplus energy demand constraint, namely: ； Minimum Remaining Energy , represented as: ； (iv) Based on the monitoring network energy queue model established in step (iii), establish data queue models at different levels, specifically including: definition For time slots Arrived inside The number of data packets for the grade data. To monitor nodes in time slots Waiting to send Queue length and transmission rate of graded data It's about transmission power. and channel state Continuous functions, Indicates monitoring node In the link The data transmission rate above, then the first The data queue length of each monitoring node at the next moment is represented as follows: ， Wherein, the initial value of the queue length =0, and These represent the high-level and low-level data queues in the time slots, respectively. The number of packet losses, and the following: ， in, express The tiered data queue in the time slot The number of packets lost; Discarded due to exceeding maximum latency requirements The maximum number of packets lost in the graded data queue; The data queue backlog length must meet the following constraints: ； A time-delay queue model was introduced to establish a time-delay queue length suitable for the scraper chain condition monitoring network. The process of updating the queue length can be represented as: ， in, It is a pre-defined constant, and ; Delay queue length Stability requirements must be met: ； The maximum latency of different data levels is expressed as... ,and: ， in, express The actual maximum delay of the grade monitoring data; express Theoretical maximum delay of graded monitoring data; (v) Based on the data queue models of different levels established in step (iv), establish a model for optimizing packet loss based on the level queues, specifically including: Each time slot Taking into account energy stability constraints, the formula is defined as follows: The priority constraint formula for the transmission of data at different levels is defined as follows: The maximum delay constraint for different levels of data is defined as follows: The transmission power constraint formula is defined as follows: The required transmission rate for different levels of data is defined as follows: The network stability requirement is defined as follows: , , and The mathematical description of the optimization problem is as follows: ， In the formula, express The delay factor for graded data can be used to control the maximum delay of different grades of data to meet the different delay requirements of high and low grade data. (vi) Based on the packet loss quantity optimization problem model established in step (v) based on the hierarchical queue, a penalty factor is introduced. Rank weight factor and delay factor An improved Lyapunov drift penalty function is constructed, specifically including: Introducing grade weighting factors Furthermore, when monitoring data anomalies occur, the monitoring data level weight factor for that node is adjusted. It will be set to the highest priority When the ranking weight factor is multiplied by the actual queue or delayed queue, the queue model... Adjusted to ,make for , and The set vector, assuming time slots The improved Lyapunov function is obtained: ,in, Indicates the first monitoring nodes The length of the virtual queue at time t; This includes information on the queue lengths of all monitoring nodes in the network, resulting in the improved Lyapunov drift as follows: ， By each time slot Greedily approaching the Lyapunov drift to a value close to zero, i.e., minimizing the Lyapunov drift. This can meet the network stability requirements; Based on the principle of drift penalty, the problem of optimizing the number of lost packets is transformed into the problem of minimizing the upper limit of the penalty function for each time slot, and a non-negative control parameter is introduced. The improved Lyapunov drift penalty function is expressed as: ， in, , and These represent the penalty factor, delay factor, and grade weight factor, respectively. when When set to zero, the improved Lyapunov drift penalty function is equivalent to the problem of minimizing Lyapunov drift; when When the values ​​are normal, the system balances queue backlog, data priority, and packet loss in the monitoring network while ensuring the stability of the energy queue and data queue. (vii) Based on the improved Lyapunov drift penalty function constructed in step (vi), a network service quality optimization algorithm based on hierarchical queue awareness is proposed. Distributed algorithms are used to optimize the sub-problems of energy management, data reception, delay guarantee, power allocation, data transmission, and queue update, so as to meet the service quality requirements of the scraper chain tension monitoring network.

2. The method for optimizing the quality of service (QoS) of a wireless sensor network for monitoring scraper chain tension according to claim 1, characterized in that, In step (i), the strain gauge is fixed to the middle of the chain link of the scraper chain, and the chain link is a circular ring chain link.

3. The method for optimizing the quality of service (QoS) of a wireless sensor network for monitoring scraper chain tension according to claim 1, characterized in that, Step (3) involves establishing a monitoring network energy queue model, introducing a virtual queue, and processing data that has exceeded the maximum latency of a node by dropping packets to meet the maximum latency requirements of the monitoring data. The virtual queue length of each monitoring node at the next moment is represented as: ， in, This is an indicator function; it takes a value of 1 when the monitoring node moves to a different level area, and 0 otherwise. Furthermore, the virtual queue needs to meet stability requirements. 。 4. The method for optimizing the quality of service (QoS) of a wireless sensor network for monitoring scraper chain tension according to claim 1, characterized in that: The problem of minimizing the long-term packet loss under constraints is transformed into a problem of minimizing drift penalty in a single time slot, in any time slot. Given feasible solutions for energy management, delay guarantees, power allocation, and data transmission, the improved Lyapunov drift penalty function in step (vi) has the following upper bound: ， In the formula, It is a finite positive constant that satisfies the following condition: 。 5. The method for optimizing the quality of service (QoS) of a wireless sensor network for monitoring scraper chain tension according to claim 1, characterized in that, In step (seven), energy balance of nodes is achieved by adjusting the duty cycle, thus optimizing energy management; the data reception rate is determined according to the operating stage of the scraper chain, thus optimizing data reception; and delay guarantee is optimized by handling packet loss of data that is about to exceed the maximum delay. Power allocation is optimized by allocating transmission power based on the level weight factor and queue length; data transmission is optimized by prioritizing the transmission of level data with higher importance factors; and queue length is optimized by updating the queue length according to the queuing dynamics formula.

6. The method for optimizing the quality of service (QoS) of a wireless sensor network for scraper chain tension monitoring according to claim 5, characterized in that, When optimizing energy management, the energy state of the node is considered. By adjusting the duty cycle This is to achieve energy balance at the nodes; when a node's energy is insufficient, reduce... Otherwise, increase To achieve energy balance at the nodes, the change in duty cycle at the next time step is expressed as: ； Nodes in their remaining energy Exceeding the minimum energy required for its operation It will run when needed; otherwise, it will remain in hibernation, waiting for energy replenishment. A fault indicator factor has been introduced. That is, when the tension monitoring data shows abnormalities, the fault indicator factor defaults to 0. At this time, the node's Automatically set to 1 to ensure timely transmission of abnormal data. Represented as: 。 7. The method for optimizing the quality of service (QoS) of a wireless sensor network for scraper chain tension monitoring according to claim 5, characterized in that, When optimizing data reception, the monitoring node is in the time slot. Maximum data reception rate at time Based on the current time slot duty cycle Maximum sampling frequency The receiving rate of high-level data is determined jointly by the scraper chain operation phase; when the monitoring node is in the scraper uplink phase, the receiving rate is expressed as... Low-level receiving rate is expressed as Conversely, when the monitoring point is in the downward phase of the scraper, the reception rate of high-level data is expressed as... Low-level receiving rate is expressed as .

8. The method for optimizing the quality of service (QoS) of a wireless sensor network for monitoring scraper chain tension according to claim 5, characterized in that, When optimizing latency guarantees, in time slots Within this process, data that is about to exceed the maximum delay is packet-dropped to ensure the validity of the monitoring data; The optimization problem can be expressed as: ； The optimal solution is: ， in, express The tiered data queue in the time slot The number of packets lost on the network As a penalty factor, As the delay factor, This is the length of the virtual queue. For delayed queues, Discarded due to exceeding maximum latency requirements The maximum number of packets lost in the graded data queue. express The actual maximum delay of the grade monitoring data, express The theoretical maximum latency of the level monitoring data; when the amount of data in the queue exceeds a set threshold, the timeliness of the monitoring data will be ensured by discarding data at the tail of the queue, in order to prevent the monitoring data from becoming invalid due to exceeding the maximum latency, by adjusting the penalty factor. Change the maximum latency for all levels of data by adjusting the latency factor. Achieve maximum delay control for high- and low-level data.

9. The method for optimizing the quality of service (QoS) of a wireless sensor network for monitoring scraper chain tension according to claim 5, characterized in that, When optimizing power allocation, in any time slot In the scraper chain tension monitoring network, there are two or fewer monitoring nodes that send data to two aggregation nodes respectively; within each time slot, each aggregation node can only receive data from one monitoring node in its area, meaning there is only one monitoring node. Allocated transmission power, the rest Each node is in sleep mode, taking into account the level weighting factor. Transmission rate and queue length The node's transmission power allocation strategy is expressed as: ， in, Indicates the aggregation node The number of monitoring nodes in the area; when monitoring data anomalies occur, the priority factor of the monitoring data collected by that node will be set to the highest priority. Meanwhile, the transmission power is set to [default value]. This ensures timely early warning of scraper chain malfunctions.

10. The method for optimizing the quality of service (QoS) of a wireless sensor network for monitoring scraper chain tension according to claim 5, characterized in that, When optimizing data transmission, utilize the actual latency. and rank weighting factor Determine the importance of the data; assume the importance factor is... ,in This serves as a data flow control factor, determined by comparing factors of varying importance. The size is used to determine the current time slot. The data type that should be sent should be prioritized, with data of higher importance factors being transmitted first; when the transmission rate... Greater than the currently transmitted Queue backlog length of graded data At that time, in the remaining channel capacity Transmit other levels of data within the channel to improve channel utilization.

Images