Mine hoisting roller stress monitoring sensor network and energy consumption optimization method thereof

Abstract

The invention discloses a mine hoisting roller stress monitoring sensor network and an energy consumption optimization method thereof, and aims to minimize the overall energy consumption of the network on the premise of ensuring different levels of data priorities of the mine hoisting roller stress monitoring network. According to the invention, the penalty factor and the grade weight factor are respectively designed by combining the operation characteristics of the mine hoisting drum and utilizing the improved Lyapunov optimization drift and penalty technology, so that the stability of the monitoring network is ensured, the balance between the network transmission energy consumption and the priorities of different grades of data is realized, and the data transmission efficiency is improved. And the problem of insufficient energy supply of the monitoring network is relieved. According to the penalty factor V designed by the invention, the system can balance and monitor the relationship between the network queue length and the transmission power on the premise of ensuring the stability of the dynamic queue and the network, so that the optimization of the transmission power is realized. The priority scheduling of different levels of data can be realized by the level weight factor wr designed by the invention, so that the priority requirements of different levels of data are met.

Description

Technical Field

[0001] The invention relates to the field of mine hoisting roller stress monitoring, and in particular to a mine hoisting roller stress monitoring sensor network and an energy consumption optimization method thereof. Background Art

[0002] The safe development and utilization of strategic mineral resources is a major strategic need for national resource security. In the process of underground strategic mineral resource development, the mine hoisting system composed of hoisting containers, drums, wire ropes, etc. is currently the only tool for transporting deep solid mineral resources. It is known as the "throat" of the mine, and its reliability is directly related to the safety of life and property. With the increasing depletion of shallow coal resources, deepening and large-scale mining have become an inevitable trend in coal mining, and the load of the hoisting system has also increased from 32t to more than 50t, which puts higher requirements on its safe operation and maintenance. Due to personnel misoperation, core equipment failure and other reasons, mine accidents caused by abnormal operation of the mine hoisting system often occur. Common mine hoisting equipment operation accidents mainly include overwinding, can jamming and can falling. In-depth analysis shows that these accidents are caused by the extra load on the hoisting container during the hoisting process, and will cause the stress on the surface of the hoisting drum to increase. Therefore, it is very necessary to monitor the stress information of the hoisting drum to realize intelligent monitoring and perception of dangerous conditions and ensure the safe operation of the mine hoisting system.

[0003] Wireless sensor networks have the advantages of flexible deployment, self-powered and high scalability, and have been widely used in the condition monitoring of electromechanical equipment in coal mines. However, there are still two problems when applying wireless sensor networks to the monitoring of hoisting roller stress: on the one hand, due to the rotation characteristics of the hoisting roller, the batteries of the monitoring network nodes are difficult to be replaced in time, and there is a problem of insufficient energy supply. On the other hand, during the operation of the hoisting roller, the stress change degree of the upper surface is more significant than that of the lower surface of the roller, which means that the priority of the upper surface monitoring data is higher than that of the lower surface, which is more conducive to accurately detecting the operating status of the hoisting roller. Therefore, how to reduce the transmission energy consumption of the monitoring network on the basis of ensuring the priority of the monitoring data, so as to extend the service life of the hoisting roller stress monitoring network system with limited energy, has become a challenge. However, existing work mainly focuses on the energy stability or service quality optimization of the monitoring network, and it is difficult to simultaneously meet the energy and service quality requirements of the wireless sensor network for mine hoisting roller stress monitoring. Summary of the invention

[0004] In view of the above-mentioned technical deficiencies, the purpose of the present invention is to provide a mine hoisting roller stress monitoring sensor network and its energy consumption optimization method, which can achieve a trade-off between network transmission energy consumption and different levels of data priority under the premise of ensuring the stability of the monitoring network, alleviate the problem of insufficient energy supply of the monitoring network, and further realize high-quality continuous monitoring of mine hoisting roller stress.

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

[0006] The present invention provides a mine hoisting drum stress monitoring sensor network and an energy consumption optimization method thereof, comprising the following steps:

[0007] Step 1: Establish a wireless sensor network system model for mine hoisting drum stress monitoring;

[0008] When the drum changes from a normal state to an abnormal state, the stress change in the area close to its upper surface is the largest, while the stress change in the lower surface area is not obvious, which means that compared with the data sampled from the lower surface of the lifting drum, the data sampled from the upper surface is more conducive to accurate fault detection. Therefore, it is necessary to classify the data levels at different positions of the lifting drum. Among them, the stress change in the upper surface area of ​​the lifting drum is the largest, so it is set as a high-level area; the stress change in the middle area is more significant, so it is set as a medium-level area; the stress change in the lower surface area is the smallest, so it is set as a low-level area. Based on this, the transmission priority classification method of monitoring data at different positions of the lifting drum can be expressed as:

[0009]

[0010] in, represents the priority of data transmission, θ represents the angle of the monitoring node on the lifting roller surface;

[0011] On the basis of the above, a wireless sensor network model for roller stress monitoring is established. The sensor nodes are installed on the surface of the lifting roller, and each node is equipped with a strain sensor and an angle detection sensor. Among them, the strain sensor is used to monitor stress changes in real time, and the angle detection sensor is used to obtain the current position information of the node, and determine the level of the sampling data accordingly. And all sensor nodes are equipped with wireless transceiver modules, which can send and receive data. It is assumed that the transmission link is reliable, that is, every time the monitoring node sends data, the aggregation node can successfully receive it without retransmission. Among them, the monitoring node transmits the tension data to the aggregation node in a single hop manner, and the aggregation node can only receive data from one monitoring node at a time.

[0012] Step 2: Establish data queue models of different levels;

[0013] In wireless sensor networks, each node needs to allocate power for data transmission. Assume that the transmission power allocation matrix of time slot t is P n (t) = {P n (t),n∈N}, where P n (t) is the link L n ∈L; Assume that the capacity buffer of each node is limited and P n (t) must satisfy the constraint: 0≤P n (t)≤P max ;

[0014] Define the link state set as {S n (t),n∈N}, where S n (t) is the channel state of the link at time slot t and satisfies independent distribution; assuming that the link capacity is C n (t)∈[0,C max ], then link L n The relationship between the upper bound of the data capacity and the allocated transmission power is expressed as C n (t)≤εP max ; Among them, the positive constant ε limits the transmission rate of the node under unit power; therefore, it is further concluded that the link L at time slot t n Data transfer rate μ n The relationship between (t) and link capacity is expressed as: ∑μ n (t)≤C n (t);

[0015] From step 1, we know that the monitoring data of the upper surface area of ​​the lifting drum has the highest priority, the monitoring data of the middle area has the second highest priority, and the monitoring data of the lower surface area has the lowest priority. Therefore, the network priority of the monitoring data should meet the following conditions:

[0016]

[0017] in, and Respectively represent the priorities of high, medium and low level data transmission; Indicates that when the monitoring data exceeds the threshold or an abnormality occurs, the priority of the node monitoring data will be set to the highest level to achieve the priority sending of abnormal data, thereby ensuring the safe operation of the lifting drum;

[0018] Transmission rate μ n (t) is the transmission power P n (t) and the channel state S n (t) is a continuous function, so μ n (P n (t),S n(t)) represents the monitoring node n on link L n The data transmission rate on the network; Assuming that the data to be sent by each tension monitoring node is stored in the form of a data queue and follows the first-in-first-out principle for data transmission, and considering that the system is discrete, that is, all data collection and transmission are performed on the time slot t∈[0,T], then define A n (t)=(A1(t),...,A N (t)) and is the number of data packets arriving at the data queue at time slot t and the length of the r-level data queue waiting to be sent; based on this, according to the level of monitoring data in the upper surface area, middle area and lower surface area of ​​the drum, the queue length of the nth monitoring node at the next moment is further expressed as:

[0019]

[0020] in, Represents the high-level data queue closest to the upper surface area, Indicates the middle-level data queue in the middle area. Indicates the low-level data queue in the lower surface area, the initial value of the queue length is 0;

[0021] In order to ensure the stability of the monitoring network, the queue length of each monitoring node must meet the following constraints:

[0022]

[0023] Step 3: Establish a transmission power optimization problem model;

[0024] In each time slot t, according to the network channel state S n (t) and the queue length information Qr(t) of different levels of data to allocate appropriate transmission power P n (t), and ensure different levels of data priority Under the premise of using a suitable transmission rate μ n (t) to meet the stability requirements of the monitoring network; that is, to minimize the long-term average transmission power under some limited constraints, including the transmission power constraint 0≤P n (t)≤P max , data rate requirement formula ∑μ n (t)≤C n (t), data transmission priority constraint Network stability requirement Thus, the following mathematical description of the optimization problem is obtained:

[0025]

[0026] Step 4: Introduce the level weight factor w r and penalty factor V, an improved Lyapunov drift penalty method is proposed;

[0027] Based on the traditional Lyapunov technique, the level weight factor w is introduced r , and set the upper surface area, middle area and lower surface area of ​​the lifting roller as high-level data weight factors w u , medium-level data weight factor w m and low-level data weight factor w d ; In addition, when the monitoring data of the monitoring node is abnormal, the weight factor w r Will automatically be set to the highest level max , to ensure the priority sending of abnormal data; based on this, a Lyapunov function based on data priority is designed, which is expressed as:

[0028]

[0029] The formula contains all the queue length information of the monitoring node, and the improved Lyapunov drift function is as follows:

[0030] △L=E{L(w r Q r (t+1)-L(w r Q r (t)))},

[0031] By continuously making the Lyapunov drift approach a value close to zero at each time slot t, that is, by minimizing the Lyapunov drift △L, the stability requirement of the monitoring network is met;

[0032] Step 5: Propose a monitoring network transmission power allocation algorithm based on data priority;

[0033] 1) Each monitoring node observes channel S n ,calculate

[0034] 2) Under the condition that the transmission power and data rate requirements are met, determine The sign of the value is The monitoring node will The value is sent to the sink node, otherwise it is not sent;

[0035] 3) The aggregation node compares all monitoring nodes in the current area value, and select the maximum The link where the value is located, and the transmission power of the monitoring node where the link is located is assigned to P n =P max, the transmission power of the remaining monitoring nodes is P n =0; when the data collected by the monitoring node is abnormal, the data level of the node will be set to the highest level w max , and the transmission power is set to P by default n =P max ; Then the transmission power allocation process of the node at any time should follow the following formula:

[0036]

[0037] 4) Update the data queue lengths of different levels of monitoring nodes

[0038] Preferably, the sensor node is equipped with a strain sensor and an angle detection sensor, wherein the strain sensor is used to monitor stress changes in real time, and the angle detection sensor is used to obtain the current position information of the node and determine the level of the sampling data accordingly.

[0039] Preferably, based on the drift penalty principle, the transmission power optimization problem of the monitoring node is converted into a minimization problem of the upper bound of the penalty function in each time slot t. After introducing a non-negative control parameter V, the Lyapunov drift penalty function based on data priority is expressed as:

[0040] △L+V∑ n E{P n (t)|w r Q r (t)},

[0041] Among them, V and w r Represent the penalty factor and level weight factor respectively, by constantly changing V and w r Adjustable P n and Q r priority; so V and w r are regarded as weight factors of transmission power and data queue level respectively; when V is set to 0, the above formula is equivalent to the Lyapunov drift minimization problem; and when V is a positive number, it can be seen that the network system makes a trade-off between queue length, data priority and transmission power.

[0042] Preferably, the transmission power optimization problem under conditional constraints is converted into a problem of minimizing the Lyapunov upper bound. In any time slot t, there is a feasible solution that satisfies the data transmission and power allocation conditions based on data priority, and the upper bound of the Lyapunov drift plus penalty is △L+V∑ n E{P n (t)|w r Q r (t)} follows the following equation:

[0043]

[0044] In the formula, B r is a finite positive constant that satisfies the following conditions:

[0045]

[0046] Therefore, the transmission power optimization problem P1 in claim 1 is transformed into the problem P2 of minimizing the Lyapunov upper bound, which can be expressed as:

[0047]

[0048] In the formula,

[0049] The beneficial effects of the present invention are:

[0050] 1. The present invention combines the operating characteristics of the mine hoisting drum and proposes a method based on Lyapunov's optimization drift plus penalty technology. While ensuring the stability of the monitoring network, it achieves a trade-off between network transmission energy consumption and the priority of different levels of data, alleviating the problem of insufficient energy supply for the monitoring network.

[0051] 2. The penalty factor V designed in the present invention enables the system to balance the relationship between the monitoring network queue length and the transmission power under the premise of ensuring the dynamic queue and network stability, thereby optimizing the transmission power.

[0052] 3. The present invention introduces the level weight factor w into the traditional Lyapunov optimization framework r , this parameter can realize the priority scheduling of data of different levels to meet the priority requirements of data of different levels.

[0053] 4. The lifting roller stress monitoring wireless sensor network and its transmission power allocation method based on data priority proposed in the present invention can be extended and applied to other electromechanical equipment health status monitoring networks with data priority requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0055] Figure 1 A flow chart of a mine hoisting drum stress monitoring sensor network and an energy consumption optimization method thereof provided by the present invention;

[0056] Figure 2A diagram of a mine hoisting system provided by the present invention;

[0057] Figure 3 The roller stress monitoring network model provided by the present invention;

[0058] Figure 4 A trend diagram of the average transmission power change of monitoring nodes under different penalty factors provided by the present invention;

[0059] Figure 5 A trend diagram of the average queue backlog length change of the monitoring node under different penalty factors provided by the present invention;

[0060] Figure 6 A trend diagram showing the average queue backlog length of the monitoring node of the present invention changing with time slots;

[0061] Figure 7 A trend diagram of key performance parameters of monitoring nodes under different high-level weight factors provided by the present invention;

[0062] Figure 8 A trend diagram of key performance parameters of monitoring nodes under different medium-level weight factors provided by the present invention;

[0063] Fig. 9 This is a trend diagram of changes in key performance parameters of monitoring nodes under different high-level and medium-level weight factors provided by the present invention. DETAILED DESCRIPTION

[0064] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0065] See also Figure 1 This embodiment provides a mine hoisting drum stress monitoring sensor network and an energy consumption optimization method thereof, which specifically includes the following steps:

[0066] Step 1: Establish a wireless sensor network system model for mine hoisting drum stress monitoring; the specific steps are as follows:

[0067] See also Figure 2In the mine hoisting system diagram provided, when the drum changes from a normal state to an abnormal state, the stress change in the area close to its upper surface is the largest, while the stress change in the lower surface area is not obvious, which means that compared with the data sampled from the lower surface of the hoisting drum, the data sampled from the upper surface is more conducive to accurate fault detection. Therefore, it is necessary to classify the data levels at different positions of the hoisting drum. Among them, the stress change in the upper surface area of ​​the hoisting drum is the largest, so it is set as a high-level area; the stress change in the middle area is more significant, and it is set as a medium-level area; the stress change in the lower surface area is the smallest, and it is set as a low-level area. Based on this, the transmission priority classification method of monitoring data at different positions of the hoisting drum can be expressed as:

[0068]

[0069] in, represents the priority of data transmission, θ represents the angle of the monitoring node on the lifting roller surface;

[0070] On the basis of the above, if Figure 3 As shown in the figure, a wireless sensor network model for roller stress monitoring is established. The sensor nodes are installed on the surface of the lifting roller, and each node is equipped with a strain sensor and an angle detection sensor. Among them, the strain sensor is used to monitor stress changes in real time, and the angle detection sensor is used to obtain the current position information of the node, and determine the level of the sampling data accordingly. And all sensor nodes are equipped with wireless transceiver modules, which can send and receive data. It is assumed that the transmission link is reliable, that is, every time the monitoring node sends data, the aggregation node can successfully receive it without retransmission. Among them, the monitoring node transmits the tension data to the aggregation node in a single hop manner, and the aggregation node can only receive data from one monitoring node at a time.

[0071] Step 2: Establish data queue models of different levels;

[0072] The specific steps are as follows:

[0073] In wireless sensor networks, each node needs to allocate power for data transmission. Assume that the transmission power allocation matrix of time slot t is P n (t) = {P n (t),n∈N}, where P n (t) is the link L n ∈L. Assume that the capacity buffer of each node is limited and P n (t) must satisfy the constraint: 0≤P n (t)≤P max ;

[0074] Define the link state set as {S n (t),n∈N}, where Sn (t) is the channel state of the link at time slot t and satisfies independent distribution. Assume that the link capacity is C n (t)∈[0,C max ], then link L n The relationship between the upper bound of the data capacity and the allocated transmission power can be expressed as C n (t)≤εP max Among them, the positive constant ε limits the transmission rate of the node under unit power. Therefore, it can be further concluded that the link L at time slot t n Data transfer rate μ n The relationship between (t) and link capacity is expressed as: ∑μ n (t)≤C n (t).

[0075] From step 1, we can see that the monitoring data of the upper surface area of ​​the lifting drum has the highest priority, the monitoring data of the middle area has the second highest priority, and the monitoring data of the lower surface area has the lowest priority. Therefore, the network priority of the monitoring data should meet the following conditions:

[0076]

[0077] in, and Respectively represent the priorities of high, medium and low level data transmission; It means that when the monitoring data exceeds the threshold or an abnormality occurs, the priority of the monitoring data of this node will be set to the highest level to achieve the priority sending of abnormal data, thereby ensuring the safe operation of the lifting drum.

[0078] Transmission rate μ n (t) is the transmission power P n (t) and the channel state S n (t) is a continuous function, so μ can be used n (P n (t),S n (t)) represents the monitoring node n on link L n Assuming that the data to be sent by each tension monitoring node is stored in the form of a data queue and follows the first-in first-out (FIFO) principle for data transmission, and considering that the system is discrete, that is, all data collection and transmission are performed on the time slot t∈[0,T], then A can be defined as n (t)=(A1(t),...,A N (t)) and is the number of data packets arriving at the data queue at time slot t and the length of the r-level data queue waiting to be sent. Based on this, according to the level of monitoring data in the upper surface area, middle area and lower surface area of ​​the drum, the queue length of the nth monitoring node at the next moment can be further expressed as:

[0079]

[0080] in, Represents the high-level data queue closest to the upper surface area, Indicates the middle-level data queue in the middle area. Indicates the low-level data queue in the lower surface area, the initial value of the queue length is 0.

[0081] In order to ensure the stability of the monitoring network, the queue length of each monitoring node must meet the following constraints:

[0082]

[0083] Step 3: Establish a transmission power optimization problem model;

[0084] The specific steps are as follows:

[0085] In each time slot t, according to the network channel state S n (t) and the queue length information Qr(t) of different levels of data to allocate appropriate transmission power P n (t), and ensure different levels of data priority Under the premise of using a suitable transmission rate μ n (t) to meet the requirements of monitoring network stability. That is, to minimize the long-term average transmission power under some limited constraints, including the transmission power constraint 0≤P n (t)≤P max , data rate requirement formula ∑μ n (t)≤C n (t), data transmission priority constraint Network stability requirement Thus, the following mathematical description of the optimization problem can be obtained:

[0086]

[0087] Step 4: Introduce the level weight factor w r and penalty factor V, an improved Lyapunov drift penalty method is proposed;

[0088] The specific steps are as follows:

[0089] Based on the traditional Lyapunov technique, the level weight factor w is introducedr , and the area closest to the upper surface of the lifting drum, the middle area, and the lower surface area are set as high-level data weight factors w u , medium-level data weight factor w m and low-level data weight factor w d In addition, when the monitoring data of the monitoring node is abnormal, the weight factor w r Will automatically be set to the highest level max , to ensure the priority sending of abnormal data. Based on this, a Lyapunov function based on data priority is designed, which is expressed as:

[0090]

[0091] The formula contains all queue length information of the monitoring node. The improved Lyapunov drift function can be further obtained as follows:

[0092] △L=E{L(w r Q r (t+1)-L(w r Q r (t)))},

[0093] By continuously making the Lyapunov drift approach a value close to zero at each time slot t, that is, by minimizing the Lyapunov drift ΔL, the stability requirement of the monitoring network can be met.

[0094] Based on the drift penalty principle, the transmission power optimization problem of the monitoring node can be converted into the minimization problem of the upper bound of the penalty function in each time slot t. After introducing a non-negative control parameter V, the Lyapunov drift penalty function based on data priority can be expressed as:

[0095] △L+V∑ n E{P n (t)|w r Q r (t)},

[0096] Among them, V and w r Represent the penalty factor and weight factor respectively. By constantly changing V and w r Adjustable P n and Q r So we can put V and w r are regarded as the weight factors of transmission power and data queue level respectively. When V is set to 0, the above formula is equivalent to the Lyapunov drift minimization problem; and when V is a positive constant, it can be seen that the network system makes a trade-off between queue length, data priority and transmission power;

[0097] The present invention converts the transmission power optimization problem under conditional constraints into the problem of minimizing the Lyapunov upper bound. In any time slot t, there is a feasible solution that satisfies the data transmission and power allocation conditions based on data priority, and the upper bound of the Lyapunov drift plus penalty is △L+V∑ n E{P n (t)|w r Q r (t)} follows the following equation:

[0098]

[0099] In the formula, B r is a finite positive constant that satisfies the following conditions:

[0100]

[0101] The proof is as follows:

[0102] Expand when μ n (t)≥0,A n When (t)≥0, the following equation holds true:

[0103]

[0104] Based on the above formula, we can expand it to:

[0105]

[0106] The proof can be obtained after sorting.

[0107] Therefore, the transmission power optimization problem P1 is transformed into the problem of minimizing the Lyapunov upper bound P2, which can be expressed as:

[0108]

[0109] In the formula,

[0110] Step 5: Propose a monitoring network transmission power allocation algorithm based on data priority;

[0111] The specific steps are as follows:

[0112] 1) Each monitoring node observes channel S n ,calculate

[0113] 2) Under the condition that the transmission power and data rate requirements are met, determine The sign of the value is The monitoring node will The value is sent to the sink node, otherwise it is not sent;

[0114] 3) The aggregation node compares all monitoring nodes in the current area value, and select the maximum The link where the value is located, and the transmission power of the monitoring node where the link is located is assigned to P n =P max , the transmission power of the remaining monitoring nodes is P n =0; when the data collected by the monitoring node is abnormal, the data level of the node will be set to the highest level w max , and the transmission power is set to P by default n =P max ; Then the transmission power allocation process of the node at any time should follow the following formula:

[0115]

[0116] 4) Update the data queue lengths of different levels of monitoring nodes

[0117] The data priority-based monitoring network transmission power allocation algorithm proposed in this paper studies different penalty factors V and level weight factors w through theoretical numerical simulation. r Impact on improving the performance of the roller stress monitoring network:

[0118] The present invention first studies the influence of different penalty factors V on network performance, specifically:

[0119] 1) The present invention analyzes the average transmission power of monitoring nodes under different penalty factors V, such as Figure 4 As shown in the figure, as the penalty factor V increases, the average transmission power of the entire network decreases and quickly converges to the optimal value. The main reason is that when the penalty factor V is small, the system only optimizes the stability of the network and has little impact on the transmission power. As the penalty factor V increases, the system not only ensures the stability of the network, but also continuously balances the average data queue length and average transmission power of the monitoring network.

[0120] 2) The present invention analyzes the average queue backlog length of monitoring nodes under different penalty factors V. Figure 5 It can be seen that a larger value of V will lead to a greater degree of system queue congestion. At the same time, the average queue length is almost linearly related to the penalty factor V, which can provide some guidance for the upper limit of the storage capacity of sensor nodes in practical applications. Figure 6The trend of the queue backlog length of the monitoring node changing with the time slot t shows that there is a certain upper limit to the queue backlog length of the monitoring node. Under different penalty factors V, the average queue backlog length of the monitoring node first increases with the increase of time slot t, and finally fluctuates around a certain deterministic value. Further analysis shows that the increase of penalty factor V can achieve the stability of the network system under large capacity, while the reduction of penalty factor V can make the network system under small capacity quickly stable.

[0121] Table 1 Parameter settings for different experimental groups

[0122]

[0123] Referring to Table 1 above, the parameter settings of different experimental groups, the present invention then studies the different level weight factors w r The impact on network performance is as follows:

[0124] 1) The present invention applies different high-level data level weight factors w u The data transmission rate is analyzed as follows: Figure 7 As shown in (a), with the high-level weight factor w u As the number of high-level data transmissions increases, the overall number of transmissions of high-level data increases, while the number of transmissions of medium-level data and low-level data decreases. This shows that adjusting the weight factor can change the importance of monitoring data at different levels, thereby ensuring the timely transmission of high-level data.

[0125] The present invention uses different high-level data level weight factors w u The average queue backlog length and average transmission power under are analyzed, such as Figure 7 As shown in (b) and (c), with the high-level weight factor w u As the queue backlog length decreases first and then increases slowly, the average transmission power of the network system continues to increase. Therefore, a high-level weight factor w can be used u To achieve a trade-off between the high-level data priority of the boost drum, queue backlog length, and network transmission power;

[0126] 2) The present invention applies different mid-level data level weight factors w m The data transmission rate is analyzed as follows: Figure 8 As shown in (a), with the intermediate weight factor w m As the number of medium-level data transmissions continues to increase, the overall number of transmissions of high-level data and low-level data shows an upward trend, while the number of transmissions of high-level data and low-level data shows a downward trend.

[0127] The present invention uses different medium-level data level weight factors w m The average queue backlog length and average transmission power under are analyzed, such as Figure 8 As shown in (b) and (c), with the intermediate weight factor w m As the queue backlog length decreases first and then increases slowly, the average transmission power of the network system continues to increase. Therefore, the medium-level weight factor w can be used. m To achieve a trade-off between the medium-level data priority of the boost drum, queue backlog length, and network transmission power;

[0128] 3) The present invention analyzes the data transmission rate under different combinations of high and medium level weight factors, such as Fig. 9 As shown in (a), with the high-level weight factor w u and the intermediate weight factor w m As the number of high-level data and medium-level data transmissions increases, the number of low-level data transmissions decreases.

[0129] The present invention analyzes the average queue backlog length and average transmission power under different high and medium level weight factor combinations, such as Fig. 9 As shown in (b) and (c), as the level weight factor increases, the queue backlog length continues to decrease, while the average transmission power of the network system continues to increase. Therefore, a high level weight factor w can be used u and the intermediate weight factor w m To achieve the trade-off between high and medium data priority of the boost roller, queue backlog length, and network transmission power.

[0130] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A mine hoisting drum stress monitoring sensor network and its energy consumption optimization method, characterized in that: The following steps are involved: Step 1: Establish a wireless sensor network system model for mine hoisting drum stress monitoring; The data transmitted by wireless sensors at different positions on the lifting drum surface are collected and classified. The sensor nodes are installed on the lifting drum surface. The monitoring nodes transmit the tension data to the sink node in a single-hop manner, and the sink node can only receive data from one monitoring node at a time. The upper surface area of ​​the lifting drum with the largest stress change is set as the high level area; the stress in the middle area of ​​the lifting drum is set as the medium level area; the lower surface area of ​​the lifting drum with the smallest stress change is set as the low level area. Based on this, the transmission priority classification method of the monitoring data at different positions of the lifting drum is expressed as follows: in, represents the priority of sending data, and θ represents the angle of the monitoring node on the surface of the lifting roller. Based on this, a wireless sensor network model for roller stress monitoring is established. Step 2: Establish data queue models of different levels; In wireless sensor networks, each node needs to allocate power for data transmission. Assume that the transmission power allocation matrix of time slot t is P n (t) = {P n (t),n∈N}, where P n (t) is the link L n ∈L; Assume that the capacity buffer of each node is limited and P n (t) must satisfy the constraint: 0≤P n (t)≤P max ; Define the link state set as {S n (t),n∈N}, where S n (t) is the channel state of the link at time slot t and satisfies independent distribution; assuming that the link capacity is C n (t)∈[0,C max ], then link L n The relationship between the upper bound of the data capacity and the allocated transmission power is expressed as C n (t)≤εP max ; Among them, the positive constant ε limits the transmission rate of the node under unit power; therefore, it is further concluded that the link L at time slot t n Data transfer rate μ n The relationship between (t) and link capacity is expressed as: ∑μ n (t)≤C n (t); From step 1, we know that the monitoring data of the upper surface area of ​​the lifting drum has the highest priority, the monitoring data of the middle area has the second highest priority, and the monitoring data of the lower surface area has the lowest priority. Therefore, the network priority of the monitoring data should meet the following conditions: in, and Respectively represent the priorities of high, medium and low level data transmission; Indicates that when the monitoring data exceeds the threshold or an abnormality occurs, the priority of the node monitoring data will be set to the highest level to achieve the priority sending of abnormal data, thereby ensuring the safe operation of the lifting drum; Transmission rate μ n (t) is the transmission power P n (t) and the channel state S n (t) is a continuous function, so μ n (P n (t),S n (t)) represents the monitoring node n on link L n The data transmission rate on the network; Assuming that the data to be sent by each tension monitoring node is stored in the form of a data queue and follows the first-in-first-out principle for data transmission, and considering that the system is discrete, that is, all data collection and transmission are performed on the time slot t∈[0,T], then define A n (t)=(A1(t),...,A N (t)) and is the number of data packets arriving at the data queue at time slot t and the length of the r-level data queue waiting to be sent; based on this, according to the level of monitoring data in the upper surface area, middle area and lower surface area of ​​the drum, the queue length of the nth monitoring node at the next moment is further expressed as: in, Represents the high-level data queue closest to the upper surface area, Indicates the middle-level data queue in the middle area. Indicates the low-level data queue in the lower surface area, the initial value of the queue length is 0; In order to ensure the stability of the monitoring network, the queue length of each monitoring node must meet the following constraints: Step 3: Establish a transmission power optimization problem model; In each time slot t, according to the network channel state S n (t) and queue length information Q of different levels of data r (t) Allocate appropriate transmission power P n (t), and ensure different levels of data priority Under the premise of using a suitable transmission rate μ n (t) to meet the stability requirements of the monitoring network; that is, to minimize the long-term average transmission power under some limited constraints, including the transmission power constraint 0≤P n (t)≤P max , data rate requirement formula ∑μ n (t)≤C n (t), data transmission priority constraint Network stability requirement Thus, the following mathematical description of the optimization problem is obtained: Step 4: Introduce the level weight factor w r and penalty factor V, an improved Lyapunov drift penalty method is proposed; Based on the traditional Lyapunov technique, the level weight factor w is introduced r , and set the upper surface area, middle area and lower surface area of ​​the lifting roller as high-level data weight factors w u , medium-level data weight factor w m and low-level data weight factor w d ; In addition, when the monitoring data of the monitoring node is abnormal, the weight factor w r Will automatically be set to the highest level max , to ensure the priority sending of abnormal data; based on this, a Lyapunov function based on data priority is designed, which is expressed as: The formula contains all the queue length information of the monitoring node, and the improved Lyapunov drift function is as follows: △L=E{L(w r Q r (t+1)-L(w r Q r (t)))}, By continuously making the Lyapunov drift approach a value close to zero at each time slot t, that is, by minimizing the Lyapunov drift △L, the stability requirement of the monitoring network is met; Step 5: Propose a monitoring network transmission power allocation algorithm based on data priority; 1) Each monitoring node observes channel S n ,calculate 2) Under the condition that the transmission power and data rate requirements are met, determine The sign of the value is The monitoring node will The value is sent to the sink node, otherwise it is not sent; 3) The aggregation node compares all monitoring nodes in the current area value, and select the maximum The link where the value is located, and the transmission power of the monitoring node where the link is located is assigned to P n =P max , the transmission power of the remaining monitoring nodes is P n =0; when the data collected by the monitoring node is abnormal, the data level of the node will be set to the highest level w max , and the transmission power is set to P by default n =P max ; Then the transmission power allocation process of the node at any time should follow the following formula: 4) Update the data queue lengths of different levels of monitoring nodes 2. A mine hoisting drum stress monitoring sensor network and its energy consumption optimization method as claimed in claim 1, characterized in that: The sensor node is equipped with a strain sensor and an angle detection sensor. The strain sensor is used to monitor stress changes in real time, and the angle detection sensor is used to obtain the current position information of the node and determine the level of the sampling data accordingly.

3. A mine hoisting drum stress monitoring sensor network and its energy consumption optimization method as claimed in claim 1, characterized in that: Based on the drift penalty principle, the transmission power optimization problem of the monitoring node is converted into the minimization problem of the upper bound of the penalty function in each time slot t. After introducing a non-negative control parameter V, the Lyapunov drift penalty function based on data priority is expressed as: Among them, V and w r Represent the penalty factor and level weight factor respectively, by constantly changing V and w r Adjustable P n and Q r priority; so V and w r are regarded as weight factors of transmission power and data queue level respectively; when V is set to 0, the above formula is equivalent to the Lyapunov drift minimization problem; and when V is a positive number, it can be seen that the network system makes a trade-off between queue length, data priority and transmission power.

4. A mine hoisting drum stress monitoring sensor network and energy consumption optimization method thereof as claimed in claim 1, characterized in that: The transmission power optimization problem under conditional constraints is converted into the problem of minimizing the Lyapunov upper bound. In any time slot t, there is a feasible solution that satisfies the data transmission and power allocation conditions based on data priority, and the upper bound of the Lyapunov drift plus penalty is △L+V∑ n E{P n (t)|w r Q r (t)} follows the following equation: In the formula, B r is a finite positive constant that satisfies the following conditions: Therefore, the transmission power optimization problem P1 in claim 1 is transformed into the problem P2 of minimizing the Lyapunov upper bound, which can be expressed as: In the formula,

Images