Sensor Access Power Determination Method, Device, Computer Equipment and Storage Medium

Abstract

The present application relates to a method, apparatus, computer device, and storage medium for determining the access power of sensors. The method includes: obtaining sensor data within a preset time period, inputting the sensor data within the preset time period into a trained correlation prediction neural network, and outputting a prediction matrix, where each element in the prediction matrix includes the occurrence probability of a corresponding emergency event type and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type; determining a target emergency event and a target sensor according to the occurrence probability of the corresponding emergency event type and each correlation coefficient; and determining the scheduling decision and access power of the target sensor through an optimal decision formula. Using this method can improve the accuracy of sensor access and reduce the energy consumption of sensor access.

Description

Technical Field

[0001] The present application relates to the technical field of sensors, and particularly to a method, device, computer device, and storage medium for determining the access power of sensors. Background Art

[0002] With the increase in power sites, in order to respond to power site emergencies, it is necessary to rely on sensors to monitor the power equipment parameters and operating environment parameters in the power sites. Currently, the sensors used for monitoring are in an unconnected state under normal conditions and are only considered to be connected in the event of an emergency.

[0003] In traditional technologies, the way sensors are connected is to manually remotely monitor whether an emergency occurs in the power site to control whether the sensors need to be connected, which has the problem of low accuracy, and the access power of the sensors cannot be dynamically adjusted, resulting in high energy consumption. Summary of the Invention

[0004] Based on this, in view of the above technical problems, it is necessary to provide a method, device, computer device, and storage medium for determining the access power of sensors that can improve the accuracy of sensor access and reduce the energy consumption of sensor access.

[0005] In a first aspect, the present application provides a method for determining the access power of sensors. The method includes:

[0006] Obtain sensor data within a preset time period, input the sensor data within the preset time period into a trained correlation prediction neural network, and output a prediction matrix. Each element in the prediction matrix includes the occurrence probability of the corresponding emergency event type, and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type;

[0007] Determine a target emergency event and a target sensor according to the occurrence probability of the corresponding emergency event type and each correlation coefficient;

[0008] Determine the scheduling decision and access power of the target sensor through an optimal decision formula.

[0009] In one embodiment, the determining the scheduling decision and access power of the target sensor through the optimal decision formula includes:

[0010] For the target sensor, calculate the decision values corresponding to all optional decisions through the optimal decision formula, select the optional decision corresponding to the maximum decision value, and determine the scheduling decision and access power of the target sensor according to the optional decision.

[0011] In one embodiment, the optimal decision formula includes:

[0012]

[0013]

[0014]

[0015]

[0016] D i,j (h) = min{F i (h) + A i (h), τC i,j (h)};

[0017]

[0018] Among them, the target emergency event is the k-th emergency event type, and the target sensor is the i-th sensor type. is the optimal decision belonging to the k-th emergency event type in the t-th time slot. is the optimal decision value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the t-th time slot, γ k,i (t) is the correlation coefficient between the k-th emergency event type and the i-th sensor type in the t-th time slot. is the decision evaluation value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the t-th time slot, h is the h-th time slot before the t-th time slot within the preset time period, l k (h) is the occurrence probability of the k-th emergency event type in the h-th time slot, x k,i (h) is used to indicate whether the i-th sensor type belonging to the k-th emergency event type accesses in the h-th time slot, y i,j (h) is used to indicate whether the i-th sensor type accesses at the j-th optional access power in the h-th time slot, γ k,i (h) is the correlation coefficient between the k-th emergency event type and the i-th sensor type in the h-th time slot, F i (h) is the data backlog generated by the i-th sensor type at the initial moment in the h-th time slot, A i (h) is the newly added data volume generated by the i-th sensor type in the h-th time slot, τ is the length of each time slot, ω i is the channel bandwidth for data transmission of the i-th sensor type, λ i is the channel gain for data transmission of the i-th sensor type, p j is the j-th optional access power, N i is the noise power for data transmission of the i-th sensor type. Used to indicate the preference for formula exploration, For the combination of the i-th sensor type belonging to the k-th emergency event type selected before the t-th time slot within a preset time period at the j-th optional access power access combination, the total number of times the combination is selected.

[0019] In one embodiment, the training process of the correlation prediction neural network includes:

[0020] Obtain historical sensor data in multiple historical preset time periods, respectively mark the historical normal sensor data and historical abnormal sensor data in the historical sensor data, and mark the emergency event type and sensor type corresponding to each historical abnormal sensor data to obtain the marked data;

[0021] Use the marked data to statistically calculate the occurrence probability of each emergency event type in each historical preset time period, and determine the correlation coefficient between each emergency event type and each sensor type to obtain the actual correlation matrix corresponding to each historical preset time period;

[0022] Input the marked data into the initial correlation prediction neural network, and output the predicted correlation matrix. Each element in the predicted correlation matrix includes the occurrence probability of the corresponding emergency event type, and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type;

[0023] Based on the difference between the actual correlation matrix and the predicted correlation matrix, train the initial correlation prediction neural network to obtain the trained correlation prediction neural network.

[0024] In one embodiment, the network parameter optimization process of the initial correlation prediction neural network includes:

[0025] Obtain the initial network parameters and the initial prediction matrix of the initial correlation prediction neural network;

[0026] Calculate the average weighted decision evaluation value and the average decision evaluation value up to the current time slot, obtain the updated correlation coefficient between each emergency event type and each sensor type according to the average weighted decision evaluation value and the average decision evaluation value, obtain the updated prediction matrix based on the updated correlation coefficient, and use the updated prediction matrix and the initial prediction matrix to update the initial network parameters to obtain the updated network parameters;

[0027] Each element in the updated prediction matrix includes the occurrence probability of the corresponding emergency event type, and the updated correlation coefficient between the corresponding emergency event type and the corresponding sensor type.

[0028] In one embodiment, obtaining the updated correlation coefficient between each emergency event type and each sensor type according to the average weighted decision evaluation value and the average decision evaluation value, obtaining the updated prediction matrix based on the updated correlation coefficient, and using the updated prediction matrix and the initial prediction matrix to update the initial network parameters to obtain the updated network parameters, including:

[0029]

[0030]

[0031]

[0032]

[0033] where is the mean of all weighted decision evaluation values of the i-th sensor type belonging to the k-th emergency event type before the t-th time slot in the historical preset time period at the j-th optional access power, is the mean of all decision evaluation values of the i-th sensor type belonging to the k-th emergency event type before the t-th time slot in the historical preset time period at the j-th optional access power, h is the h-th time slot before the t-th time slot in the preset time period, J is the number of all optional access powers, is the optimal decision value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the h-th time slot, is the decision evaluation value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the h-th time slot, y i,j (h) is used to indicate whether the i-th sensor type accesses at the j-th optional access power in the h-th time slot, is the updated correlation coefficient between the k-th emergency event type and the i-th sensor type in the t-th time slot, ω(t + 1) is the updated network parameter, ω(t) is the initial network parameter corresponding to the t-th time slot, ρ is the step factor, ζ(t) is the initial prediction matrix corresponding to the t-th time slot, ζ Ideal (t) is the updated prediction matrix.

[0034] Second, the present application also provides a sensor access power determination device. The device includes:

[0035] A model processing module, configured to obtain sensor data within a preset time period, input the sensor data within the preset time period into a trained correlation prediction neural network, and output a prediction matrix, where each element in the prediction matrix includes the occurrence probability of a corresponding emergency event type and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type;

[0036] A target determination module, configured to determine a target emergency event and a target sensor according to the occurrence probability of the corresponding emergency event type and each correlation coefficient;

[0037] A power determination module, configured to determine the scheduling decision and access power of the target sensor through an optimal decision formula.

[0038] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of the method in any of the above embodiments are implemented.

[0039] In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method in any of the above embodiments are implemented.

[0040] In a fifth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the steps of the method in any of the above embodiments are implemented.

[0041] The above-mentioned method, device, computer equipment, and storage medium for determining the access power of sensors obtain sensor data within a preset time period, input the sensor data within the preset time period into a trained correlation prediction neural network, and output a prediction matrix. Each element in the prediction matrix includes the occurrence probability of the corresponding emergency event type and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type. The target emergency event and the target sensor are determined based on the occurrence probability of the corresponding emergency event type and each correlation coefficient. The target access power of the target sensor is determined through the power optimal decision formula. Compared with the problems of low accuracy of sensor access and high energy consumption caused by manually remotely monitoring whether an emergency event occurs in a power site to control whether the sensor needs to be accessed in the traditional technology, in this application, the trained correlation prediction neural network processes the sensor data to obtain the occurrence probability of each emergency event type and the correlation coefficient between each emergency event type and each sensor type, so as to determine the target sensor, making the determined target sensor highly accurate, improving the accuracy of target sensor access, and determining the target access power of the target sensor through the power optimal decision formula. The target access power of the target sensor is determined through the power optimal decision formula, making the determined target access power more accurate, and further achieving the reduction of the energy consumption of sensor access. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic flowchart of the method for determining the access power of sensors provided in an embodiment of the present application;

[0043] Figure 2 It is a structural block diagram of the correlation prediction neural network provided in an embodiment;

[0044] Figure 3 It is a schematic flowchart of the training process of the correlation prediction neural network in an embodiment;

[0045] Figure 4 It is a structural block diagram of a sensor access power determination device provided in an embodiment of the present application;

[0046] Figure 5 It is an internal structure diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] In order to make the objectives, technical solutions, and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0048] In this embodiment, a method for determining the access power of sensors is provided. This embodiment takes the application of this method to a computer device as an example for illustration. It can be understood that this method can also be applied to a server, and can also be applied to a system including a computer device and a server, and is implemented through the interaction between the computer device and the server.

[0049] Before describing the specific embodiments of the present application, first introduce the application scenario of the method for determining the access power of sensors. This method is applied to a power site, and the power equipment parameters and operating environment parameters of the power site need to be monitored by sensors. The sensors can use MEMS sensors, and the sensor types can be gas sensors, pressure sensors, radio frequency sensors, infrared sensors, and temperature sensors, which are not specifically limited. The sensor data collected by the sensors can be transmitted to the monitoring terminal through communication methods such as micro-power wireless, RS485, or power line carrier. The monitoring terminal stores a trained correlation prediction neural network, and the trained correlation prediction neural network is used to process the sensor data in the monitoring terminal to determine the target emergency event and the target sensor, and to determine the target access power of the target sensor through the power optimal decision formula.

[0050] Figure 1 It is a schematic flowchart of the method for determining the access power of sensors provided in the embodiments of the present application. This method is applied to a computer device or a server. In one embodiment, as Figure 1 shown, it includes the following steps:

[0051] S101, obtain the sensor data within a preset time period, input the sensor data within the preset time period into the trained correlation prediction neural network, and output a prediction matrix. Each element in the prediction matrix includes the occurrence probability of the corresponding emergency event type, and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type.

[0052] In this embodiment, there are T time slots in the preset time period, and at most one emergency event occurs in each time slot. Continuously collect sensor data within the preset time period, and input the collected sensor data into the trained correlation prediction neural network. The structural block diagram of the correlation prediction neural network is as Figure 2 shown. The correlation prediction neural network includes an input layer, a hidden layer, and an output layer. After the collected sensor data is input into the trained correlation prediction neural network at the input layer, the output layer outputs a prediction matrix. Taking there being K emergency event types and I sensor types as an example, the prediction matrix obtained in the t-th time slot within the preset time period is as follows:

[0053]

[0054] Wherein, ζ(t) is the prediction matrix, l k (t) is the occurrence probability of the k-th emergency event type in the t-th time slot, γ k,i (t) is the correlation coefficient between the k-th emergency event type and the i-th sensor type in the t-th time slot.

[0055] S102. Determine the target emergency event and the target sensor according to the occurrence probability of the corresponding emergency event type and each correlation coefficient.

[0056] In one embodiment, the method for determining the target emergency event and the target sensor according to the occurrence probability of the corresponding emergency event type and each correlation coefficient may be to select the emergency event type corresponding to the maximum occurrence probability among the occurrence probabilities of each emergency event type as the target emergency event, and determine the sensor type with the maximum correlation coefficient value with the target emergency event as the target sensor according to each correlation coefficient.

[0057] In this embodiment, through the prediction matrix output by the trained correlation prediction neural network, the emergency event type to which the target emergency event belongs can be determined more accurately, and the sensor type with the highest correlation with the target emergency event is selected as the target sensor, improving the accuracy of determining the target sensor.

[0058] S103. Determine the scheduling decision and access power of the target sensor through the optimal decision formula.

[0059] In this embodiment, for the target sensor and the target emergency event, the optimal decision value of all optional decisions is calculated through the optimal decision formula. The optional access power can be set manually. Here, there are J optional access powers set manually. Select the optional decision corresponding to the maximum value among each optimal decision value, and obtain the access decision and access power of the target sensor according to this decision. The access decision and access power of the target sensor determined by the method in this embodiment can ensure the energy utilization rate and reduce the energy consumption.

[0060] In some embodiments, the optimal decision formula includes:

[0061]

[0062]

[0063]

[0064]

[0065] D i,j (h) = min{F i (h) + A i (h), τCi,j (h)};

[0066]

[0067] where the target emergency event is the k-th emergency event type, and the target sensor is the i-th sensor type, is the optimal decision belonging to the k-th emergency event type in the t-th time slot, is the optimal decision value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the t-th time slot, γ k,i (t) is the correlation coefficient between the k-th emergency event type and the i-th sensor type in the t-th time slot, is the decision evaluation value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the t-th time slot, h is the h-th time slot before the t-th time slot within the preset time period, l k (h) is the occurrence probability of the k-th emergency event type in the h-th time slot, x k,i (h) is used to indicate whether the i-th sensor type belonging to the k-th emergency event type accesses in the h-th time slot, y i,j (h) is used to indicate whether the i-th sensor type accesses at the j-th optional access power in the h-th time slot, γ k,i (h) is the correlation coefficient between the k-th emergency event type and the i-th sensor type in the h-th time slot, F i (h) is the data backlog generated by the i-th sensor type at the initial moment in the h-th time slot, A i (h) is the newly generated data volume of the i-th sensor type in the h-th time slot, τ is the length of each time slot, ω i is the channel bandwidth for data transmission of the i-th sensor type, λ i is the channel gain for data transmission of the i-th sensor type, p j is the j-th optional access power, N i is the noise power for data transmission of the i-th sensor type, is used to indicate the preference for exploration of the formula, is the total number of times the combination of selecting the i-th sensor type belonging to the k-th emergency event type to access at the j-th optional access power before the t-th time slot within the preset time period is selected.

[0068] Further, x k,i (h) is used to indicate whether the i-th sensor type belonging to the k-th emergency event type accesses in the h-th time slot. In the case that the i-th sensor type belonging to the k-th emergency event type accesses in the h-th time slot, xk,i (h) has a value of 1. When the i-th sensor type belonging to the k-th emergency event type is not connected in the h-th time slot, x k,i (h) has a value of 0; y i,j (h) is used to indicate whether the i-th sensor type accesses with the j-th optional access power in the h-th time slot. When the i-th sensor type accesses with the j-th optional access power in the h-th time slot, y i,j (h) has a value of 1. When the i-th sensor type does not access with the j-th optional access power in the h-th time slot, y i,j (h) has a value of 0; It is used to indicate the preference for exploration in the formula. Among them, and the larger the value, the more inclined to explore, and the smaller the value, the more inclined to exploit.

[0069] In the sensor access power determination method provided in this embodiment, by obtaining sensor data within a preset time period, inputting the sensor data within the preset time period into a trained correlation prediction neural network, and outputting a prediction matrix. Each element in the prediction matrix includes the occurrence probability of the corresponding emergency event type and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type. The target emergency event and the target sensor are determined according to the occurrence probability of the corresponding emergency event type and each correlation coefficient. Through the optimal decision formula, the access decision and access power of the target sensor are determined. Compared with the problems of low accuracy of sensor access and high energy consumption caused by manually remotely monitoring whether an emergency occurs in a power site to control whether a sensor needs to access in the traditional technology, in this embodiment, the trained correlation prediction neural network processes the sensor data to obtain the occurrence probability of each emergency event type and the correlation coefficient between each emergency event type and each sensor type, so as to determine the target sensor, making the determined target sensor highly accurate, improving the accuracy of target sensor access, and determining the target access power of the target sensor through the power optimal decision formula. The target access power of the target sensor is determined by the power optimal decision formula, making the determined target access power more accurate and further achieving the reduction of the energy consumption of sensor access.

[0070] In one embodiment, the flow diagram of the training process of the correlation prediction neural network is as Figure 3 shown and includes the following steps:

[0071] S301, Obtain historical sensor data within multiple historical preset time periods, respectively mark the historical normal sensor data and historical abnormal sensor data in the historical sensor data, and mark the emergency event type and sensor type corresponding to each historical abnormal sensor data to obtain the marked data.

[0072] In this embodiment, the method of respectively marking the historical normal sensor data and the historical abnormal sensor data in the historical sensor data is to mark the historical normal sensor data as ZC and the historical abnormal sensor data as YC. The method of marking the emergency event type and the sensor type corresponding to each historical abnormal sensor data is (YC1, k, i), and (YC1, k, i) is used to represent that the first historical abnormal sensor data corresponds to the k-th emergency event type and the i-th sensor type. Taking the historical abnormal sensor data as fire data as an example, the corresponding emergency event type is fire and the sensor type is a temperature sensor.

[0073] S302. Using the marked data, statistically calculate the occurrence probability of each emergency event type within each historical preset time period, and determine the correlation coefficient between each emergency event type and each sensor type, so as to obtain the actual correlation matrix corresponding to each historical preset time period.

[0074] Taking a specific historical preset time period as an example, this historical preset time period includes 10 time slots. The first emergency event type occurred 2 times, and the second emergency event type occurred 5 times. Then the occurrence probability of the first emergency event type is 2 / 10 = 0.2. The first emergency event type occurred in the 3rd time slot, and the first emergency event type corresponds to the second sensor type, then γ 1,2 (3)=1.

[0075] S303. Input the marked data into the initial correlation prediction neural network, and output the predicted correlation matrix. Each element in the predicted correlation matrix includes the occurrence probability of the corresponding emergency event type and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type.

[0076] S304. Based on the difference between the actual correlation matrix and the predicted correlation matrix, train the initial correlation prediction neural network to obtain the trained correlation prediction neural network.

[0077] By training the initial correlation prediction neural network, the output result of the trained correlation prediction neural network can be made more accurate.

[0078] In some embodiments, the training process of the initial correlation prediction neural network is also a process of optimizing the network parameters of the initial correlation prediction neural network. The process of optimizing the network parameters of the initial correlation prediction neural network is to obtain the initial network parameters and the initial prediction matrix of the initial correlation prediction neural network; calculate the average weighted decision evaluation value and the average decision evaluation value up to the current time slot, obtain the updated correlation coefficient between each emergency event type and each sensor type according to the average weighted decision evaluation value and the average decision evaluation value, obtain the updated prediction matrix based on the updated correlation coefficient, and use the updated prediction matrix and the initial prediction matrix to update the initial network parameters to obtain the updated network parameters;

[0079] Among them, the process of obtaining the updated correlation coefficient between each emergency event type and each sensor type according to the average weighted decision evaluation value and the average decision evaluation value, obtaining the updated prediction matrix based on the updated correlation coefficient, and using the updated prediction matrix and the initial prediction matrix to update the initial network parameters to obtain the updated network parameters is implemented by the following formula:

[0080]

[0081]

[0082]

[0083]

[0084] Among them, is the mean value of all weighted decision evaluation values of the i-th sensor type belonging to the k-th emergency event type before the t-th time slot in the historical preset time period at the j-th optional access power, is the mean value of all decision evaluation values of the i-th sensor type belonging to the k-th emergency event type before the t-th time slot in the historical preset time period at the j-th optional access power. h is the h-th time slot before the t-th time slot in the preset time period, and J is the number of all optional access powers, is the optimal decision value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the h-th time slot, is the decision evaluation value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the h-th time slot, y i,j (h) is used to indicate whether the i-th sensor type accesses at the j-th optional access power in the h-th time slot, is the updated correlation coefficient between the k-th emergency event type and the i-th sensor type in the t-th time slot, ω(t + 1) is the updated network parameter, ω(t) is the initial network parameter corresponding to the t-th time slot, ρ is the step size factor, ζ(t) is the initial prediction matrix corresponding to the t-th time slot, and ζ Ideal (t) is the updated prediction matrix.

[0085] In this embodiment, optimizing the network parameters of the initial correlation prediction neural network can make the output result of the trained correlation prediction neural network more accurate.

[0086] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are shown sequentially according to the arrows, these steps do not necessarily have to be executed in the order indicated by the arrows. Unless otherwise clearly stated in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least some of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages do not necessarily have to be executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least some of the steps or steps in other steps.

[0087] Based on the same inventive concept, the embodiments of the present application also provide a sensor access power determination device for implementing the sensor access power determination method described above. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the sensor access power determination device provided below can refer to the limitations on the sensor access power determination method in the above text, and will not be repeated here.

[0088] See Figure 4 , Figure 4 is the structural block diagram of a sensor access power determination device provided in the embodiments of the present application. The device 400 includes: a model processing module 401, a target determination module 402, and a power determination module 403, where:

[0089] The model processing module 401 is configured to obtain sensor data within a preset time period, input the sensor data within the preset time period into the trained correlation prediction neural network, and output a prediction matrix. Each element in the prediction matrix includes the occurrence probability of the corresponding emergency event type, and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type;

[0090] A target determination module 402, configured to determine a target emergency event and a target sensor according to the occurrence probability of the corresponding emergency event type and each correlation coefficient.

[0091] A power determination module 403, configured to determine the scheduling decision and access power of the target sensor through an optimal decision formula.

[0092] In the sensor access power determination device provided in this embodiment, the model processing module obtains sensor data within a preset time period, inputs the sensor data within the preset time period into the trained correlation prediction neural network, and outputs a prediction matrix. Each element in the prediction matrix includes the occurrence probability of the corresponding emergency event type and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type. The target determination module determines the target emergency event and the target sensor according to the occurrence probability of the corresponding emergency event type and each correlation coefficient, and the power determination module determines the target access power of the target sensor based on the power optimal decision formula. Compared with the problems of low accuracy of sensor access and high energy consumption caused by manually remotely monitoring whether an emergency event occurs in a power site to control whether a sensor needs to be accessed in the traditional technology, in this embodiment, the trained correlation prediction neural network processes the sensor data to obtain the occurrence probability of each emergency event type and the correlation coefficient between each emergency event type and each sensor type, so as to determine the target sensor, making the determined target sensor highly accurate, improving the accuracy of target sensor access, and determining the target access power of the target sensor through the power optimal decision formula. The target access power of the target sensor is determined by the power optimal decision formula, making the determined target access power more accurate, and further realizing the reduction of the energy consumption of sensor access.

[0093] Optionally, the power determination module 403 includes:

[0094] A power determination unit, configured to calculate the optimal decision value corresponding to all optional access powers for the target sensor through the power optimal decision formula, and select the optional access power corresponding to the maximum value among the optimal decision values as the target access power of the target sensor.

[0095] Optionally, the power optimal decision formula includes:

[0096]

[0097]

[0098]

[0099]

[0100] Di,j (h) = min{F i (h)+A i (h),τC i,j (h)};

[0101]

[0102] Among them, the target emergency event is the kth emergency event type, the target sensor is the ith sensor type, is the optimal decision for the kth emergency type in the tth time slot, is the optimal decision value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the t-th time slot, γ k,i (t) is the correlation coefficient between the kth emergency event type and the ith sensor type in the tth time slot, is the decision evaluation value of the i-th sensor type belonging to the k-th emergency event type in the t-th time slot under the j-th optional access power, h is the h-th time slot before the t-th time slot in the preset time period, l k (h) is the probability of occurrence of the kth emergency event type in the hth time slot, x k,i (h) is used to indicate whether the i-th sensor type belonging to the k-th emergency event type is connected in the h-th time slot, y i,j (h) is used to indicate whether the i-th sensor type is accessed with the j-th optional access power in the h-th time slot, γ k,i (h) is the correlation coefficient between the kth emergency event type and the ith sensor type in the hth time slot, F i (h) is the data backlog generated by the i-th sensor type at the initial time in the h-th time slot, A i (h) is the amount of new data generated by the i-th sensor type in the h-th time slot, τ is the length of each time slot, ω i is the channel bandwidth for data transmission of the i-th sensor type, λ i is the channel gain for data transmission of the i-th sensor type, p j is the jth optional access power, N i is the noise power of the data transmission of the i-th sensor type, Used to indicate the formula's preference for exploration, The total number of times the i-th sensor type belonging to the k-th emergency event type is selected in the j-th optional access power access combination before the t-th time slot within the preset time period.

[0103] Optionally, the device 400 further includes:

[0104] A data marking module, configured to obtain historical sensor data within multiple historical preset time periods, mark the historical normal sensor data and historical abnormal sensor data in the historical sensor data respectively, and mark the emergency event type and sensor type corresponding to each historical abnormal sensor data, so as to obtain marked data;

[0105] A matrix obtaining module, configured to use the marked data to statistically calculate the occurrence probability of each emergency event type within each historical preset time period, and determine the correlation coefficient between each emergency event type and each sensor type, so as to obtain an actual prediction matrix corresponding to each historical preset time period;

[0106] An input / output module, configured to input the marked data into an initial correlation prediction neural network and output an initial prediction matrix, where each element in the initial prediction matrix includes the occurrence probability of the corresponding emergency event type, and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type;

[0107] A network training module, configured to train the initial correlation prediction neural network based on the difference between the actual prediction matrix and the initial prediction matrix, so as to obtain a trained correlation prediction neural network.

[0108] Optionally, the apparatus 400 further includes:

[0109] An initial parameter obtaining module, configured to obtain the initial network parameters of the initial correlation prediction neural network and the initial prediction matrix;

[0110] A parameter updating module, configured to calculate the average weighted decision evaluation value and the average decision evaluation value up to the current time slot, obtain the updated correlation coefficient between each emergency event type and each sensor type according to the average weighted decision evaluation value and the average decision evaluation value, obtain an updated prediction matrix based on the updated correlation coefficient, and update the initial network parameters by using the updated prediction matrix and the initial prediction matrix, so as to obtain updated network parameters; each element in the updated prediction matrix includes the occurrence probability of the corresponding emergency event type, and the updated correlation coefficient between the corresponding emergency event type and the corresponding sensor type.

[0111] Optionally, the parameter updating module includes:

[0112]

[0113]

[0114]

[0115]

[0116] Wherein, is the mean of all weighted decision evaluation values of the i-th sensor type belonging to the k-th emergency event type before the t-th time slot within the historical preset time period at the j-th optional access power. is the mean of all decision evaluation values of the i-th sensor type belonging to the k-th emergency event type before the t-th time slot within the historical preset time period at the j-th optional access power, h is the h-th time slot before the t-th time slot within the preset time period, J is the number of all optional access powers. is the optimal decision value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the h-th time slot. is the decision evaluation value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the h-th time slot, y i,j (h) is used to indicate whether the i-th sensor type accesses at the j-th optional access power in the h-th time slot. is the updated correlation coefficient between the k-th emergency event type and the i-th sensor type in the t-th time slot, ω(t + 1) is the updated network parameter, ω(t) is the initial network parameter corresponding to the t-th time slot, ρ is the step factor, ζ(t) is the initial prediction matrix corresponding to the t-th time slot, ζ Ideal (t) is the updated prediction matrix.

[0117] Each module in the above sensor access power determination device can be implemented in whole or in part by software, hardware, and their combination. Each of the above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0118] In one embodiment, a computer device is provided. The computer device can be a terminal, and its internal structure diagram can be as Figure 5As shown in the figure. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface, the display unit, and the input device are connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner, and the wireless manner can be implemented through 5G communication, WIFI, mobile cellular network, NFC (Near Field Communication), or other technologies. The computer program, when executed by the processor, implements a method for determining the access power of sensors. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covered on the display screen, or a button, a trackball, or a touchpad set on the shell of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0119] Those skilled in the art can understand that Figure 5 the structure shown in the figure is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0120] In one embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, it implements the steps of the method for determining the access power of sensors provided in the above embodiment:

[0121] Obtain sensor data within a preset time period, input the sensor data within the preset time period into the trained correlation prediction neural network, and output a prediction matrix. Each element in the prediction matrix includes the occurrence probability of the corresponding emergency event type, and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type;

[0122] Determine the target emergency event and the target sensor according to the occurrence probability of the corresponding emergency event type and each correlation coefficient;

[0123] Determine the access decision and access power of the target sensor through the optimal decision formula.

[0124] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0125] For the target sensor, calculate the decision values corresponding to all optional decisions through the optimal decision formula, select the optional decision corresponding to the maximum decision value, and obtain the access decision and access power of the target sensor according to this decision.

[0126] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0127]

[0128]

[0129]

[0130]

[0131] D i,j (h) = min{F i (h) + A i (h), τC i,j (h)};

[0132]

[0133] Wherein, the target emergency event is the k-th emergency event type, and the target sensor is the i-th sensor type. is the optimal decision belonging to the k-th emergency event type in the t-th time slot. is the decision value of the i-th sensor type belonging to the k-th emergency event type in the t-th time slot under the j-th optional access power, γ k,i (t) is the correlation coefficient between the k-th emergency event type and the i-th sensor type in the t-th time slot. is the decision evaluation value of the i-th sensor type belonging to the k-th emergency event type in the t-th time slot under the j-th optional access power, h is the h-th time slot before the t-th time slot within the preset time period, l k (h) is the occurrence probability of the k-th emergency event type in the h-th time slot, x k,i (h) is used to indicate whether the i-th sensor type belonging to the k-th emergency event type accesses in the h-th time slot, y i,j (h) is used to indicate whether the i-th sensor type accesses with the j-th optional access power in the h-th time slot, γ k,i (h) is the correlation coefficient between the k-th emergency event type and the i-th sensor type in the h-th time slot, F i(h) is the amount of data backlog generated by the i-th sensor type at the initial moment in the h-th time slot, A i (h) is the amount of new data generated by the i-th sensor type in the h-th time slot, τ is the length of each time slot, ω i is the channel bandwidth for data transmission of the i-th sensor type, λ i is the channel gain for data transmission of the i-th sensor type, p j is the j-th optional access power, N i is the noise power for data transmission of the i-th sensor type, used to indicate the preference of the formula for exploration, is the total number of times the combination of the i-th sensor type belonging to the k-th emergency event type accessing at the j-th optional access power is selected before the t-th time slot within a preset time period.

[0134] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0135] Obtain historical sensor data within multiple historical preset time periods, respectively mark the historical normal sensor data and historical abnormal sensor data in the historical sensor data, and mark the emergency event type and sensor type corresponding to each historical abnormal sensor data to obtain the marked data;

[0136] Use the marked data to statistically calculate the occurrence probability of each emergency event type within each historical preset time period, and determine the correlation coefficient between each emergency event type and each sensor type to obtain the actual correlation matrix corresponding to each historical preset time period;

[0137] Input the marked data into the initial correlation prediction neural network, and output the prediction correlation matrix. Each element in the initial prediction matrix includes the occurrence probability of the corresponding emergency event type, and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type;

[0138] Based on the difference between the actual correlation matrix and the prediction correlation matrix, train the initial correlation prediction neural network to obtain the trained correlation prediction neural network.

[0139] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0140] Obtain the initial network parameters and the initial prediction matrix of the initial correlation prediction neural network;

[0141] Calculate the average weighted decision evaluation value and the average decision evaluation value up to the current time slot, obtain the updated correlation coefficient between each emergency event type and each sensor type according to the average weighted decision evaluation value and the average decision evaluation value, obtain the updated prediction matrix based on the updated correlation coefficient, and update the initial network parameters by using the updated prediction matrix and the initial prediction matrix to obtain the updated network parameters;

[0142] Each element in the updated prediction matrix includes the occurrence probability of the corresponding emergency event type and the updated correlation coefficient between the corresponding emergency event type and the corresponding sensor type.

[0143] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0144]

[0145]

[0146]

[0147]

[0148] where, is the mean value of all weighted decision evaluation values of the i-th sensor type belonging to the k-th emergency event type before the t-th time slot in the historical preset time period at the j-th optional access power, is the mean value of all decision evaluation values of the i-th sensor type belonging to the k-th emergency event type before the t-th time slot in the historical preset time period at the j-th optional access power, h is the h-th time slot before the t-th time slot in the preset time period, J is the number of all optional access powers, is the optimal decision value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the h-th time slot, is the decision evaluation value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the h-th time slot, y i,j (h) is used to indicate whether the i-th sensor type accesses at the j-th optional access power in the h-th time slot, is the updated correlation coefficient between the k-th emergency event type and the i-th sensor type in the t-th time slot, ω(t + 1) is the updated network parameter, ω(t) is the initial network parameter corresponding to the t-th time slot, ρ is the step factor, ζ(t) is the initial prediction matrix corresponding to the t-th time slot, ζ Ideal (t) is the updated prediction matrix.

[0149] The implementation principles and technical effects of the above embodiments are similar to those of the above method embodiments, and will not be elaborated here.

[0150] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the sensor access power determination method provided in the above embodiment are implemented:

[0151] Obtain sensor data within a preset time period, input the sensor data within the preset time period into a trained correlation prediction neural network, and output a prediction matrix. Each element in the prediction matrix includes the occurrence probability of the corresponding emergency event type and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type;

[0152] Determine a target emergency event and a target sensor according to the occurrence probability of the corresponding emergency event type and each correlation coefficient;

[0153] Determine the access decision and access power of the target sensor through an optimal decision formula.

[0154] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0155] For the target sensor, calculate the decision values corresponding to all optional decisions through the optimal decision formula, select the optional decision corresponding to the maximum decision value, and obtain the access decision and access power of the target sensor based on this decision.

[0156] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0157]

[0158]

[0159]

[0160]

[0161] D i,j (h) = min{F i (h) + A i (h), τC i,j (h)};

[0162]

[0163] Wherein, the target emergency event is the k-th emergency event type, and the target sensor is the i-th sensor type, is the optimal decision belonging to the k-th emergency event type in the t-th time slot, The decision value of the $i$-th sensor type belonging to the $k$-th emergency event type at the $j$-th optional access power in the $t$-th time slot, $\gamma$ k,i $\rho_{ik}(t)$ is the correlation coefficient between the $k$-th emergency event type and the $i$-th sensor type in the $t$-th time slot, The decision evaluation value of the $i$-th sensor type belonging to the $k$-th emergency event type at the $j$-th optional access power in the $t$-th time slot, $h$ is the $h$-th time slot before the $t$-th time slot within a preset time period, $l$ k $P_k(h)$ is the occurrence probability of the $k$-th emergency event type in the $h$-th time slot, $x$ k,i $x_{ik}(h)$ is used to indicate whether the $i$-th sensor type belonging to the $k$-th emergency event type accesses in the $h$-th time slot, $y$ i,j $y_{ij}(h)$ is used to indicate whether the $i$-th sensor type accesses with the $j$-th optional access power in the $h$-th time slot, $\gamma$ k,i $\rho_{ik}(h)$ is the correlation coefficient between the $k$-th emergency event type and the $i$-th sensor type in the $h$-th time slot, $F$ i $A_i(h)$ is the data backlog generated by the $i$-th sensor type at the initial moment in the $h$-th time slot, $A$ i $B_i(h)$ is the newly generated data volume of the $i$-th sensor type in the $h$-th time slot, $\tau$ is the length of each time slot, $\omega$ i $W_i$ is the channel bandwidth for data transmission of the $i$-th sensor type, $\lambda$ i $G_i$ is the channel gain for data transmission of the $i$-th sensor type, $p$ j $P_j$ is the $j$-th optional access power, $N$ i $N_i$ is the noise power for data transmission of the $i$-th sensor type, $\epsilon$ is used to indicate the preference for exploration of the formula, $C_{ijk}$ is the total number of times the combination of selecting the $i$-th sensor type belonging to the $k$-th emergency event type to access with the $j$-th optional access power is selected before the $t$-th time slot within a preset time period.

[0164] In one embodiment, when the computer program is executed by the processor, the following steps are further implemented:

[0165] Obtain historical sensor data in multiple historical preset time periods, respectively mark the historical normal sensor data and historical abnormal sensor data in the historical sensor data, and mark the emergency event type and sensor type corresponding to each historical abnormal sensor data to obtain the marked data;

[0166] Use the marked data to statistically calculate the occurrence probability of each emergency event type in each historical preset time period, and determine the correlation coefficient between each emergency event type and each sensor type to obtain the actual correlation matrix corresponding to each historical preset time period;

[0167] Input the marked data into the initial correlation prediction neural network to output a predicted correlation matrix, where each element in the predicted correlation matrix includes the occurrence probability of the corresponding emergency event type and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type.

[0168] Train the initial correlation prediction neural network based on the difference between the actual correlation matrix and the predicted correlation matrix to obtain a trained correlation prediction neural network.

[0169] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0170] Obtain the initial network parameters and the initial prediction matrix of the initial correlation prediction neural network.

[0171] Calculate the average weighted decision evaluation value and the average decision evaluation value up to the current time slot, obtain the updated correlation coefficient between each emergency event type and each sensor type according to the average weighted decision evaluation value and the average decision evaluation value, obtain an updated prediction matrix based on the updated correlation coefficient, and update the initial network parameters using the updated prediction matrix and the initial prediction matrix to obtain updated network parameters.

[0172] Each element in the updated prediction matrix includes the occurrence probability of the corresponding emergency event type and the updated correlation coefficient between the corresponding emergency event type and the corresponding sensor type.

[0173] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0174]

[0175]

[0176]

[0177]

[0178] Among them, is the mean value of all weighted decision evaluation values of the i-th sensor type belonging to the k-th emergency event type before the t-th time slot at the j-th optional access power within the historical preset time period. is the mean value of all decision evaluation values of the i-th sensor type belonging to the k-th emergency event type before the t-th time slot at the j-th optional access power within the historical preset time period, h is the h-th time slot before the t-th time slot within the preset time period, and J is the number of all optional access powers. The optimal decision value for the $i$-th sensor type belonging to the $k$-th emergency event type at the $j$-th optional access power in the $h$-th time slot The decision evaluation value $y$ for the $i$-th sensor type belonging to the $k$-th emergency event type at the $j$-th optional access power in the $h$-th time slot i,j $\tau_{ij}^{(h)}$ is used to indicate whether the $i$-th sensor type accesses at the $j$-th optional access power in the $h$-th time slot $\omega^{(t + 1)}$ is the updated correlation coefficient between the $k$-th emergency event type and the $i$-th sensor type in the $t$-th time slot, $\omega^{(t)}$ is the initial network parameter corresponding to the $t$-th time slot, $\rho$ is the step size factor, $\zeta^{(t)}$ is the initial prediction matrix corresponding to the $t$-th time slot, and $\zeta^{(t)}$ Ideal (t) is the updated prediction matrix

[0179] The implementation principle and technical effects of the above embodiments are similar to those of the above method embodiments, and will not be elaborated here

[0180] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of the sensor access power determination method provided in the above embodiments:

[0181] Obtain sensor data within a preset time period, input the sensor data within the preset time period into the trained correlation prediction neural network, and output a prediction matrix. Each element in the prediction matrix includes the occurrence probability of the corresponding emergency event type and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type

[0182] Determine the target emergency event and the target sensor according to the occurrence probability of the corresponding emergency event type and each correlation coefficient

[0183] Determine the access decision and access power of the target sensor through the optimal decision formula

[0184] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0185] For the target sensor, calculate the decision values of all optional decisions through the optimal decision formula, select the optional decision corresponding to the maximum decision value, and obtain the access decision and access power of the target sensor according to this decision

[0186] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0187]

[0188]

[0189]

[0190]

[0191] D i,j (h) = min{F i (h) + A i (h), τC i,j (h)};

[0192]

[0193] Among them, the target emergency event is the k-th emergency event type, and the target sensor is the i-th sensor type. is the optimal decision belonging to the k-th emergency event type in the t-th time slot. is the decision value of the i-th sensor type belonging to the k-th emergency event type in the t-th time slot under the j-th optional access power, γ k,i (t) is the correlation coefficient between the k-th emergency event type and the i-th sensor type in the t-th time slot. is the decision evaluation value of the i-th sensor type belonging to the k-th emergency event type in the t-th time slot under the j-th optional access power. h is the h-th time slot before the t-th time slot within the preset time period, l k (h) is the occurrence probability of the k-th emergency event type in the h-th time slot, x k,i (h) is used to indicate whether the i-th sensor type belonging to the k-th emergency event type accesses in the h-th time slot, y i,j (h) is used to indicate whether the i-th sensor type accesses with the j-th optional access power in the h-th time slot, γ k,i (h) is the correlation coefficient between the k-th emergency event type and the i-th sensor type in the h-th time slot, F i (h) is the data backlog generated by the i-th sensor type at the initial moment in the h-th time slot, A i (h) is the newly added data volume generated by the i-th sensor type in the h-th time slot. τ is the length of each time slot, ω i is the channel bandwidth for data transmission of the i-th sensor type, λ i is the channel gain for data transmission of the i-th sensor type, p j is the j-th optional access power, N i is the noise power for data transmission of the i-th sensor type. is used to indicate the preference for exploration of the formula. The total number of times a combination of the i-th sensor type belonging to the k-th emergency event type is selected at the j-th optional access power access combination before the t-th time slot within a preset time period.

[0194] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0195] Obtain historical sensor data within multiple historical preset time periods, respectively mark the historical normal sensor data and historical abnormal sensor data in the historical sensor data, and mark the emergency event type and sensor type corresponding to each historical abnormal sensor data to obtain the marked data;

[0196] Use the marked data to statistically calculate the occurrence probability of each emergency event type within each historical preset time period, and determine the correlation coefficient between each emergency event type and each sensor type to obtain the actual correlation matrix corresponding to each historical preset time period;

[0197] Input the marked data into the initial correlation prediction neural network, and output the prediction correlation matrix. Each element in the prediction correlation matrix includes the occurrence probability of the corresponding emergency event type, and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type;

[0198] Based on the difference between the actual correlation matrix and the prediction correlation matrix, train the initial correlation prediction neural network to obtain the trained correlation prediction neural network.

[0199] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0200] Obtain the initial network parameters and the initial prediction matrix of the initial correlation prediction neural network;

[0201] Calculate the average weighted decision evaluation value and the average decision evaluation value up to the current time slot, obtain the updated correlation coefficient between each emergency event type and each sensor type according to the average weighted decision evaluation value and the average decision evaluation value, obtain the updated prediction matrix based on the updated correlation coefficient, and use the updated prediction matrix and the initial prediction matrix to update the initial network parameters to obtain the updated network parameters;

[0202] Each element in the updated prediction matrix includes the occurrence probability of the corresponding emergency event type, and the updated correlation coefficient between the corresponding emergency event type and the corresponding sensor type.

[0203] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0204]

[0205]

[0206]

[0207]

[0208] Among them, is the mean value of all weighted decision evaluation values of the i-th sensor type belonging to the k-th emergency event type before the t-th time slot in the historical preset time period at the j-th optional access power. is the mean value of all decision evaluation values of the i-th sensor type belonging to the k-th emergency event type before the t-th time slot in the historical preset time period at the j-th optional access power. h is the h-th time slot before the t-th time slot in the preset time period, and J is the number of all optional access powers. is the optimal decision value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the h-th time slot. is the decision evaluation value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the h-th time slot, y i,j (h) is used to indicate whether the i-th sensor type accesses at the j-th optional access power in the h-th time slot. is the updated correlation coefficient between the k-th emergency event type and the i-th sensor type in the t-th time slot. ω(t + 1) is the updated network parameter, ω(t) is the initial network parameter corresponding to the t-th time slot, ρ is the step factor, ζ(t) is the initial prediction matrix corresponding to the t-th time slot, ζ Ideal (t) is the updated prediction matrix.

[0209] The implementation principle and technical effect of the above embodiment are similar to those of the above method embodiment, and will not be elaborated here.

[0210] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memory can include Read-Only Memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., without limitation.

[0211] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as within the scope described in this specification.

[0212] The above-described embodiments merely represent several implementation manners of the present application. Their descriptions are relatively specific and detailed, but they should not be construed as limiting the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A method for determining the access power of a sensor, characterized in that The method includes: Obtain sensor data within a preset time period, input the sensor data within the preset time period into a trained correlation prediction neural network, and output a prediction matrix. Each element in the prediction matrix includes the occurrence probability of the corresponding emergency event type and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type; Determine a target emergency event and a target sensor according to the occurrence probability of the corresponding emergency event type and each correlation coefficient; For the target sensor, calculate the decision values corresponding to all optional decisions through an optimal decision formula, select the optional decision corresponding to the maximum decision value, and determine the scheduling decision and access power of the target sensor according to the optional decision; The optimal decision formula includes: ； ； ； ； ； ； where the target emergency event is the k-th emergency event type, and the target sensor is the i-th sensor type, is the optimal decision belonging to the k-th emergency event type in the t-th time slot, is the optimal decision value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the t-th time slot, is the correlation coefficient between the k-th emergency event type and the i-th sensor type in the t-th time slot, is the decision evaluation value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the t-th time slot, h is the h-th time slot before the t-th time slot within the preset time period, is the occurrence probability of the k-th emergency event type in the h-th time slot, is used to indicate whether the i-th sensor type belonging to the k-th emergency event type accesses in the h-th time slot, is used to indicate whether the i-th sensor type accesses at the j-th optional access power in the h-th time slot, is the correlation coefficient between the k-th emergency event type and the i-th sensor type in the h-th time slot, is the data backlog generated by the i-th sensor type at the initial moment in the h-th time slot, is the newly generated data volume of the i-th sensor type in the h-th time slot, is the length of each time slot, is the channel bandwidth for data transmission of the i-th sensor type, is the channel gain for data transmission of the i-th sensor type, is the j-th optional access power, is the noise power of data transmission of the i-th sensor type, is used to indicate the preference of the formula for exploration, is the total number of times the combination of the i-th sensor type belonging to the k-th emergency event type accessing at the j-th optional access power is selected before the t-th time slot within the preset time period.

2. The method according to claim 1, characterized in that, The training process of the correlation prediction neural network includes: Obtain historical sensor data within multiple historical preset time periods, respectively mark the historical normal sensor data and historical abnormal sensor data in the historical sensor data, and mark the emergency event type and sensor type corresponding to each historical abnormal sensor data to obtain marked data; Use the marked data to statistically calculate the occurrence probability of each emergency event type within each historical preset time period, and determine the correlation coefficient between each emergency event type and each sensor type to obtain an actual correlation matrix corresponding to each historical preset time period; Input the marked data into an initial correlation prediction neural network, and output a predicted correlation matrix. Each element in the predicted correlation matrix includes the occurrence probability of the corresponding emergency event type and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type; Based on the difference between the actual correlation matrix and the predicted correlation matrix, train the initial correlation prediction neural network to obtain a trained correlation prediction neural network.

3. The method according to claim 2, wherein The network parameter optimization process of the initial correlation prediction neural network includes: Obtain the initial network parameters and the initial prediction matrix of the initial correlation prediction neural network; Calculate the average weighted decision evaluation value and the average decision evaluation value up to the current time slot, obtain the updated correlation coefficient between each emergency event type and each sensor type according to the average weighted decision evaluation value and the average decision evaluation value, obtain an updated prediction matrix based on the updated correlation coefficient, and use the updated prediction matrix and the initial prediction matrix to update the initial network parameters to obtain updated network parameters; Each element in the updated prediction matrix includes the occurrence probability of the corresponding emergency event type and the updated correlation coefficient between the corresponding emergency event type and the corresponding sensor type.

4. The method according to claim 3, wherein The updated correlation coefficient between each emergency event type and each sensor type is obtained according to the average weighted decision evaluation value and the average decision evaluation value, and an updated prediction matrix is obtained based on the updated correlation coefficient. The initial network parameters are updated by using the updated prediction matrix and the initial prediction matrix to obtain updated network parameters, including: ； ； ； ； wherein, is the mean of all weighted decision evaluation values of the i-th sensor type belonging to the k-th emergency event type before the t-th time slot within the historical preset time period at the j-th optional access power, is the mean of all decision evaluation values of the i-th sensor type belonging to the k-th emergency event type before the t-th time slot within the historical preset time period at the j-th optional access power, h is the h-th time slot before the t-th time slot within the preset time period, and J is the number of all optional access powers, is the optimal decision value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the h-th time slot, is the decision evaluation value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the h-th time slot, is used to indicate whether the i-th sensor type accesses at the j-th optional access power in the h-th time slot, is the updated correlation coefficient between the k-th emergency event type and the i-th sensor type in the t-th time slot, is the updated network parameter, is the initial network parameter corresponding to the t-th time slot, is the step factor, is the initial prediction matrix corresponding to the t-th time slot, is the updated prediction matrix.

5. A sensor access power determination device, characterized in that, The device includes: A model processing module, configured to obtain sensor data within a preset time period, input the sensor data within the preset time period into a trained correlation prediction neural network, and output a prediction matrix, where each element in the prediction matrix includes the occurrence probability of a corresponding emergency event type and the correlation coefficient between the corresponding emergency event type and the corresponding sensor type; A target determination module, configured to determine a target emergency event and a target sensor according to the occurrence probability of the corresponding emergency event type and each correlation coefficient; A power determination module, configured to, for the target sensor, calculate decision values corresponding to all optional decisions through an optimal decision formula, select the optional decision corresponding to the maximum decision value, and determine the scheduling decision and access power of the target sensor according to the optional decision; The optimal decision formula includes: ； ； ； ； ； ； where the target emergency event is the k-th emergency event type, and the target sensor is the i-th sensor type, is the optimal decision belonging to the k-th emergency event type in the t-th time slot, is the optimal decision value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the t-th time slot, is the correlation coefficient between the k-th emergency event type and the i-th sensor type in the t-th time slot, is the decision evaluation value of the i-th sensor type belonging to the k-th emergency event type at the j-th optional access power in the t-th time slot, h is the h-th time slot before the t-th time slot within the preset time period, is the occurrence probability of the k-th emergency event type in the h-th time slot, is used to indicate whether the i-th sensor type belonging to the k-th emergency event type accesses in the h-th time slot, is used to indicate whether the i-th sensor type accesses at the j-th optional access power in the h-th time slot, is the correlation coefficient between the k-th emergency event type and the i-th sensor type in the h-th time slot, is the data backlog generated by the i-th sensor type at the initial moment in the h-th time slot, is the newly generated data volume of the i-th sensor type in the h-th time slot, is the length of each time slot, is the channel bandwidth for data transmission of the i-th sensor type, is the channel gain for data transmission of the i-th sensor type, is the j-th optional access power, is the noise power for data transmission of the i-th sensor type, is used to indicate the preference for exploration of the formula, is the total number of times the combination of selecting the i-th sensor type belonging to the k-th emergency event type to access at the j-th optional access power is selected before the t-th time slot within the preset time period.

6. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1 to 4 are implemented.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 4 are implemented.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 4 are implemented.

Images