Safety warning method and device for limited space, electronic equipment and storage medium

Abstract

The application discloses a limited space safety early warning method and device, electronic equipment and storage medium, and relates to the technical field of electric power. The method comprises the following steps: determining an actual safety score in a limited space according to target data in the limited space, wherein the target data at least comprises environment data, equipment data, personnel state data of operating personnel and operation data in the limited space; inputting the actual safety score into a safety state neural network to obtain a target score, wherein the target score is used for representing an overall safety risk level in the limited space; and generating early warning information according to a warning threshold, the actual safety score and the target score, wherein the warning threshold is periodically adjusted according to the actual safety score and the target score. The application solves the technical problem that the existing technology often causes poor early warning reliability due to safety early warning of the limited space based on a single data source.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric power, in particular to a safety warning method and device for limited space, electronic equipment and storage medium. BACKGROUND

[0002] In electric power construction, limited space operation often involves closed or poorly ventilated areas, including shafts of electric power equipment, tunnels, interiors of substations, etc. There are many challenges when operating in these environments, therefore, it is crucial to ensure the safety of operating personnel in limited space.

[0003] However, current warning systems rely on a single data source, such as monitoring of gas concentration or oxygen level. This method has certain limitations, as a single data source cannot comprehensively reflect all potential risks in limited space. In addition, due to the rapid and highly dynamic changes in the environment within limited space, such dynamic changes cannot be captured in real time by a single data source, which affects the accuracy of the warning. Furthermore, the physiological state of operating personnel is an important factor affecting the safety of operation, but existing warning systems usually ignore this point. Moreover, current warning systems cannot adjust the warning method in real time according to the state of operating personnel and the changes in the environment within limited space, resulting in unreliable warnings.

[0004] To address the above problems, no effective solutions have been proposed so far. SUMMARY

[0005] The present application provides a safety warning method and device for limited space, electronic equipment and storage medium, to at least solve the technical problem of poor warning reliability caused by safety warning of limited space based on a single data source in the prior art.

[0006] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a safety warning method for limited space is provided, comprising: determining an actual safety score in the limited space according to target data in the limited space, wherein the target data at least includes environmental data, equipment data, personnel state data of operating personnel, and operation data in the limited space; inputting the actual safety score into a safety state neural network to obtain a target score, wherein the target score is used to represent the overall safety risk level in the limited space; generating warning information according to a warning threshold, the actual safety score and the target score, wherein the warning threshold is periodically adjusted according to the actual safety score and the target score.

[0007] Optionally, the safety state neural network is trained by the following steps, comprising: obtaining historical data in the limited space, wherein the historical data at least includes historical environmental data and historical personnel state data of the operating personnel in the limited space; extracting key features from the historical data, and fusing the key features according to the weight of each key feature to obtain a feature vector at different time, wherein the key features at least include historical environmental parameters and physiological characteristics of the operating personnel in the limited space, wherein the historical environmental parameters at least include temperature, humidity, gas concentration, and oxygen content, and the physiological characteristics of the operating personnel at least include heart rate, blood pressure, and blood oxygen saturation; obtaining a multi-modal data set according to the feature vector and a risk score value of each feature vector; and training the safety state neural network based on the multi-modal data set.

[0008] Optionally, the safety state neural network is trained based on the multi-modal data set, comprising: obtaining an initial learning rate, an input layer neuron number, and an output layer neuron number of an initial neural network, wherein the initial learning rate is used to represent the update frequency of the initial setting of the weight and bias vector in the initial neural network; determining a target learning rate of the initial neural network according to the initial learning rate and a decay factor, wherein the target learning rate is used to improve the convergence speed of the initial neural network, and the decay factor is used to represent the rate of decay of the initial learning rate; determining a target number of hidden layers and a neuron number of each hidden layer of the initial neural network according to the input layer neuron number and the output layer neuron number, wherein the hidden layer is located between the input layer and the output layer of the initial neural network, and the target number and the neuron number are used to improve the performance of the initial neural network; and performing multiple iteration training on the initial neural network according to the multi-modal data set to obtain the safety state neural network.

[0009] Optionally, after the safety state neural network is trained based on the multi-modal data set, the safety warning method for the limited space further comprises: dividing the multi-modal data set into N subsets, wherein N is an integer greater than 1; performing K rounds of test operations on the safety state neural network to obtain K prediction results, wherein each round of test operation is used to select any one of the N subsets as a target test set, and the subsets other than the target test set are used as a target training set, the safety state neural network is tested for one round using the target training set, and the target test set is predicted using the tested safety state neural network to obtain one prediction result; determining K accuracies of the safety state neural network according to the K prediction results, wherein the Jth accuracy in the K accuracies is used to represent the prediction accuracy of the safety state neural network in the Jth test operation in the K rounds of test operations; and calculating the average of the K accuracies as the prediction accuracy of the safety state neural network.

[0010] Optionally, the generating the early warning information according to the early warning threshold, the actual safety score and the target score comprises: calculating an absolute difference between the actual safety score and the target score; updating the early warning threshold according to the absolute difference and an adjustment coefficient to obtain a target early warning threshold; and generating the early warning information according to the target early warning threshold and the actual safety score.

[0011] Optionally, the generating the early warning information according to the target early warning threshold and the actual safety score comprises: determining a first value according to the environmental data in the limited space and a weight of the environmental data, wherein the first value is used to represent a risk intensity caused by an environmental factor in the limited space; taking an absolute difference between the first value and a second value as a target difference; determining a risk type in the limited space according to the target early warning threshold, the first value, the second value and the target difference, wherein the risk type is used to distinguish a source category of the limited space risk; determining a risk level in the limited space according to the actual safety score and at least one preset risk threshold, wherein the risk level is used to represent a risk degree in the limited space; and generating the early warning information according to the risk type and the risk level.

[0012] Optionally, the determining the risk type in the limited space according to the target early warning threshold, the first value, the second value and the target difference comprises: in a case that the first value is greater than the second value and the target difference is greater than the target early warning threshold, determining that a first type of risk occurs in the limited space, wherein the first type of risk is used to represent a risk caused by an environmental factor; in a case that the first value is less than the second value and the target difference is greater than the target early warning threshold, determining that a second type of risk occurs in the limited space, wherein the second type of risk is used to represent a risk caused by a physical factor of an operating personnel; and in a case that the target difference is less than or equal to the target early warning threshold, determining that a third type of risk occurs in the limited space, wherein the third type of risk is used to represent a risk caused by the environmental factor and the physical factor of the operating personnel.

[0013] According to another aspect of the present application, a safety early warning device for a limited space is also provided, comprising: a determination unit configured to determine an actual safety score in the limited space according to target data in the limited space, wherein the target data at least comprises environmental data, equipment data, personnel state data of an operating personnel and operation data in the limited space; an input unit configured to input the actual safety score into a safety state neural network to obtain a target score, wherein the target score is used to represent an overall safety risk level in the limited space; and a generation unit configured to generate early warning information according to an early warning threshold, the actual safety score and the target score, wherein the early warning threshold is periodically adjusted according to the actual safety score and the target score.

[0014] According to another aspect of the present application, an electronic device is also provided, wherein the electronic device comprises one or more processors and a memory, and the memory is configured to store one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors are caused to perform the safety warning method for a limited space.

[0015] According to another aspect of the present application, a computer readable storage medium is also provided, wherein the computer readable storage medium stores a computer program, and when the computer program is executed, the device where the computer readable storage medium is located is caused to perform the safety warning method for a limited space.

[0016] From the above, in the present application, firstly, the actual safety score in the limited space is determined according to a plurality of data sources in the limited space, which guarantees the comprehensiveness of the data sources, then the actual safety score is input into the safety state neural network to obtain a target score, wherein the target score is used to represent the overall safety risk level in the limited space, and then the warning information is generated according to the warning threshold, the actual safety score and the target score, wherein the warning threshold is periodically adjusted according to the actual safety score and the target score, which ensures the reliability of the warning.

[0017] Compared with the safety warning for the work personnel working in the limited space based on a single data source in the prior art, the actual safety score of the limited space is determined according to the real-time environment data, equipment data, state data of the work personnel and work data in the limited space, the actual safety score value determined according to the plurality of data sources is input into the safety state neural network to obtain a target score, then the warning threshold in the limited space is periodically adjusted according to the actual safety score obtained from the real-time data and the target score obtained from the safety state neural network, which achieves the purpose of adjusting the warning threshold in real time according to the real-time data in the limited space, and generates the warning information for reminding, thereby realizing the technical effect of improving the reliability of the warning, and further solving the technical problem that the safety warning for the limited space is often based on a single data source in the prior art, which leads to poor reliability of the warning. BRIEF DESCRIPTION OF DRAWINGS

[0018] The accompanying drawings, which are included to provide a further understanding of the present application, constitute a part of the present application and illustrate the illustrative embodiments of the present application and their description serve to explain the present application, and do not constitute improper limitations on the present application. In the drawings:

[0019] Figure 1 is a flowchart of an optional safety warning method for a limited space according to an embodiment of the present application;

[0020] Figure 2is a schematic view of an optional safety warning device for a limited space according to an embodiment of the present application. DETAILED DESCRIPTION

[0021] In order to make the personnel in the technical field better understand the scheme of the present application, the technical scheme in the embodiments of the present application will be clearly and completely described below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative labor should belong to the protection scope of the present application.

[0022] It should be noted that the terms "first", "second", and the like in the specification and claims of the present application and the above-described drawings are used to distinguish similar objects, and do not necessarily have to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0023] It should also be noted that the information and data collected by the present application are information and data authorized by the user or authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of related data comply with relevant laws, regulations and standards in relevant regions, necessary security measures are taken, do not violate public order and good customs, and provide corresponding operation portal for user to choose authorization or refusal. For example, the system and the interface between the related users or institutions are provided, and before obtaining the related information, the interface needs to send an acquisition request to the aforementioned user or institution, and after receiving the consent information fed back by the aforementioned user or institution, the related information is acquired.

[0024] According to an embodiment of the present application, an embodiment of a safety warning method for a limited space is provided. It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described herein can be executed in a different order from that shown herein.

[0025] It should be noted that an intelligent early warning system can serve as an execution subject of the safety early warning method for the limited space according to the embodiments of the present application. It can be understood that the safety early warning method for the limited space provided by the embodiments of the present application can also be executed by other systems or devices as the execution subject, which is not specifically limited by the embodiments of the present application.

[0026] Figure 1 is a flowchart of an optional safety early warning method for the limited space according to the embodiments of the present application, as shown in the figure, the method comprises the following steps: Figure 1

[0027] Step S101, determining an actual safety score in the limited space according to target data in the limited space.

[0028] In step S101, the target data at least includes environmental data, equipment data, personnel state data of the operating personnel, and operation data in the limited space.

[0029] Optionally, the environmental data in the limited space includes but is not limited to temperature, humidity, gas concentration, and oxygen content in the limited space.

[0030] Optionally, the personnel state data of the operating personnel in the limited space includes but is not limited to heart rate, blood pressure, and blood oxygen saturation of the operating personnel.

[0031] Step S102, inputting the actual safety score into a safety state neural network to obtain a target score.

[0032] In step S102, the target score is used to represent the overall safety risk level in the limited space.

[0033] Optionally, the intelligent early warning system inputs the actual safety score into the trained safety state neural network to obtain a predicted score value in the limited space.

[0034] Step S103, generating early warning information according to a warning threshold, the actual safety score, and the target score.

[0035] In step S103, the warning threshold is periodically adjusted according to the actual safety score and the target score.

[0036] Optionally, the warning threshold is adjusted in real time according to the actual safety score and the target score.

[0037] ​It can be known from the contents of steps S101 to S103 that, in the present application, first, the actual safety score in the limited space is determined according to the target data in the limited space, wherein the target data at least includes environmental data, equipment data, personnel state data of the operating personnel, and operation data in the limited space, then the actual safety score is input into the safety state neural network to obtain a target score, wherein the target score is used to represent the overall safety risk level in the limited space, and finally, the warning information is generated according to the warning threshold, the actual safety score and the target score, wherein the warning threshold is periodically adjusted according to the actual safety score and the target score.

[0038] It can be known from the above contents that, in the present application, first, the actual safety score in the limited space is determined according to a plurality of data sources in the limited space, which guarantees the comprehensiveness of the data sources, then the actual safety score is input into the safety state neural network to obtain a target score, wherein the target score is used to represent the overall safety risk level in the limited space, and then the warning information is generated according to the warning threshold, the actual safety score and the target score, wherein the warning threshold is periodically adjusted according to the actual safety score and the target score, which ensures the reliability of the warning.

[0039] Compared with the safety warning for the operating personnel in the limited space based on a single data source in the prior art, the present application determines the actual safety score of the limited space according to the real-time environmental data, equipment data, state data of the operating personnel and operation data in the limited space, inputs the actual safety score value determined according to a plurality of data sources into the safety state neural network to obtain a target score, then periodically adjusts the warning threshold in the limited space according to the actual safety score obtained from the real-time data and the target score obtained from the safety state neural network, achieves the purpose of real-time adjustment of the warning threshold according to the real-time data in the limited space, and generates warning information for reminding, thereby realizing the technical effect of improving the reliability of the warning, and further solving the technical problem that the safety warning for the limited space is often based on a single data source in the prior art, which leads to poor reliability of the warning.

[0040] In an optional embodiment, the intelligent early warning system first acquires historical data in the limited space, wherein the historical data at least includes historical environmental data and historical personnel state data of the operating personnel in the limited space, then extracts key features from the historical data, fuses the key features according to the weight of each key feature to obtain a feature vector at different time, wherein the key features at least include historical environmental parameters and physiological characteristics of the operating personnel in the limited space, wherein the historical environmental parameters at least include temperature, humidity, gas concentration and oxygen content, and the physiological characteristics of the operating personnel at least include heart rate, blood pressure and blood oxygen saturation, then obtains a multi-modal data set according to the feature vector and the risk score value of each feature vector, and finally trains a safety state neural network based on the multi-modal data set.

[0041] Optionally, the intelligent early warning system first acquires historical data in the limited space, and extracts key features from the historical data, wherein the key features in the historical environmental data in the limited space at least include temperature, humidity, gas concentration and oxygen content, and the key features in the personnel state data of the operating personnel at least include heart rate, blood pressure and blood oxygen saturation of the operating personnel, extracts these key features, fuses the key features to obtain a fused feature vector, obtains different feature vectors according to different samples, then obtains a multi-modal data set according to each feature vector and the risk score value corresponding to each feature vector, and finally trains a safety state neural network based on the multi-modal data set.

[0042] Optionally, the intelligent early warning system processes the historical environmental data, calculates the mean value T kz of all temperatures collected in the historical time period kz , and calculates the change rate of the gas concentration G in a given time interval, as shown in formula (1):

[0043]

[0044] Wherein, ΔG(t) represents the change rate of the gas concentration G, G(t) represents the gas concentration G at the current time point t, G(t-Δt) represents the gas concentration G corresponding to the time point t-Δt, and Δt represents the time interval (seconds).

[0045] Optionally, the intelligent early warning system calculates the change rate of the oxygen content OC in a given time interval, as shown in formula (2):

[0046]

[0047] Wherein, ΔOC(t) represents the change rate of the oxygen content OC, OC(t) represents the oxygen content OC at the current time point t, and OC(t-Δt) represents the oxygen content OC corresponding to the time point t-Δt.

[0048] Optionally, the intelligent early warning system processes the personnel state data of the historical workers, calculates the mean value BP of the blood pressure of each worker collected in the historical time period kz , and calculates the heart rate variation coefficient of each worker for evaluating the amplitude of the heart rate variation of each worker, as shown in formula (3):

[0049]

[0050] wherein, HRV represents the heart rate variation coefficient of each worker, HR i represents the i-th heart rate measurement value of each worker, HR kz represents the mean value of all heart rate measurement values of each worker, and n represents the number of heart rate measurement values of each worker.

[0051] Optionally, the intelligent early warning system calculates the change rate of the blood oxygen saturation of each worker in a given time interval, as shown in formula (4):

[0052]

[0053] wherein, ΔSP(t) represents the change rate of the blood oxygen saturation SP of each worker, SP(t) represents the blood oxygen saturation SP of each worker at the current time point t, and SP(t-Δt) represents the blood oxygen saturation SP of each worker at the time point t-Δt.

[0054] Optionally, the intelligent early warning system fuses the extracted features, as shown in formula (5):

[0055] F tud = [w1·ΔG(t) + w2·ΔOC(t) + w3·HRV + w4·ΔSP(t) + w5·T kz +w6·H kz +

[0056] w7·BP kz ](5)

[0057] wherein, F tud is the fused feature vector, w1, w2, w3, w4, w5, w6 and w7 respectively represent the weights of ΔG(t), ΔOC(t), HRV, ΔSP(t), T kz , H kz and BP kz .

[0058] Optionally, the intelligent early warning system obtains a multi-modal data set based on the fused features and the risk score values of each feature vector, as shown in formula (6):

[0059] D = {(F1, S1), (F2, S2), (F3, S3),..., (Fu, Su)} (6) u , Su) (6) u )} (6)

[0060] wherein, D is a multi-modal data set, Fu is a feature vector after fusion of the u-th sample, Su represents a risk score value after fusion of the u-th sample, and u is the number of samples. u u

[0061] Optionally, the safety state neural network refers to a BP (Backpropagation Neural Network Model) neural network model.

[0062] Optionally, the BP neural network model comprises an input layer, a hidden layer and an output layer, wherein the input layer comprises seven nodes corresponding to the fused D respectively, the hidden layer comprises 10 nodes, and the output layer comprises one node, the input is an actual safety score S acl , and the output is a comprehensive safety score S zs (target score), wherein the actual safety score S acl is preset by a manager based on environmental data, equipment data, personnel state data of operating personnel and operation data in a limited space.

[0063] Optionally, the intelligent early warning system initializes the weight matrix W and the bias vector b of the BP neural network model, as shown in formula (6):

[0064]

[0065] wherein, n inc is the number of input nodes, 0 represents a mean value, indicating that the expected value of the weight matrix W is 0, represents a standard deviation, indicating that the standard deviation of the weight matrix W is determined by n inc , and indicates sampling in a normal distribution with a mean value of 0 and a variance of .

[0066] Optionally, the intelligent early warning system adopts forward propagation to calculate the score prediction value S pv of the output layer, as shown in formulas (7)-(9):

[0067] Z (1) = W (1) D + b (1) (7)

[0068] A (1) = f(Z (1) ) (8)​​

[0069] S pv = W (2) A (1) + b (2) (9)

[0070] where Z (1) represents the linear combination of the hidden layer, A (1) represents the activation output of the hidden layer, W (1) and b (1) are the weight matrix and bias vector input to the hidden layer, respectively, W (2) and b (2) are the weight matrix and bias vector from the hidden layer to the output layer, respectively, f(Z (1) ) represents the activation function of the hidden layer.

[0071] Optionally, the intelligent early warning system calculates the difference between the score prediction value S pv and the comprehensive safety score S zs through the mean square error, as shown in equation (10):

[0072]

[0073] where L represents the loss function between the score prediction value S pv and the comprehensive safety score S zs .

[0074] Optionally, the intelligent early warning system calculates the gradient and updates the weight matrix and bias vector by backpropagation, and calculates the gradient of the output layer, as shown in equation (11):

[0075]

[0076] where δ (2) is the error term of the output layer, is the gradient of the loss function L with respect to the score prediction value S pv .

[0077] Optionally, the intelligent early warning system calculates the gradient of the hidden layer, as shown in equation (12):

[0078] δ (1) = δ (2) W (2) f d (Z (1) ) (12)

[0079] where δ (1) is the error term of the hidden layer, and f d (Z (1) ) is the derivative of the activation function f of the hidden layer.

[0080] Optionally, the intelligent early warning system calculates the gradients of the weight matrix and the bias, as shown in equations (13)-(16):

[0081]

[0082] where, is the gradient of the loss function L with respect to the weight matrix W (2) from the hidden layer to the output layer, is the gradient of the loss function L with respect to the bias vector b (2) from the hidden layer to the output layer, is the gradient of the loss function L with respect to the weight matrix W (1) from the input layer to the hidden layer, is the gradient of the loss function L with respect to the bias vector b (1) from the input layer to the hidden layer.

[0083] Optionally, the intelligent early warning system updates the weight matrix and the bias vector, as shown in equations (17) and (18):

[0084]

[0085] where, W (i) is the weight matrix of the i-th layer, b (i) is the bias vector of the i-th layer, and η is the learning rate, which controls the step size of parameter updates.

[0086] From the above, it can be seen that the intelligent early warning system integrates historical environmental data and physiological state data of workers in limited space, providing a comprehensive information base. By extracting key features, the system can identify key factors affecting safety, such as environmental parameters (temperature, humidity, gas concentration, oxygen content) and physiological characteristics (heart rate, blood pressure, blood oxygen saturation), and the multi-modal data set constructed provides a rich source of information for system simulation and prediction, which helps to more comprehensively understand the safety situation of limited space work. Based on the trained neural network, the system can predict and warn potential safety risks in limited space, so as to take measures in advance and reduce the possibility of accidents.

[0087] In an optional embodiment, the intelligent early warning system obtains an initial learning rate, an input layer neuron number, and an output layer neuron number of an initial neural network, wherein the initial learning rate is used to represent an initial setting update frequency of a weight and a bias vector in the initial neural network, then a target learning rate of the initial neural network is determined according to the initial learning rate and a decay factor, wherein the target learning rate is used to improve a convergence speed of the initial neural network, and the decay factor is used to represent a rate of decay of the initial learning rate, then a target number of hidden layers and a neuron number of each hidden layer of the initial neural network are determined according to the input layer neuron number and the output layer neuron number, wherein the hidden layers are located between an input layer and an output layer of the initial neural network, and the target number and the neuron number are used to improve a performance of the initial neural network, and finally the initial neural network is trained multiple times according to a multi-modal data set to obtain a safety state neural network.

[0088] Optionally, the intelligent early warning system optimizes initial parameters of the BP neural network, and makes the learning rate η adaptive in a decay manner, as shown in formula (19):

[0089]

[0090] wherein η0 is the initial learning rate, ω is a current iteration number, τ is a time constant of the learning rate decay, and e is a base number of an exponential function.

[0091] Optionally, the intelligent early warning system determines an optimal number of hidden layers through cross-validation, as shown in formula (20):

[0092]

[0093] wherein Q is the number of hidden layers, is the number of input layer neurons, is the number of output layer neurons.

[0094] Optionally, the intelligent early warning system determines a configuration of neurons, as shown in formula (21):

[0095]

[0096] wherein B represents the number of neurons of each hidden layer.

[0097] Optionally, the intelligent early warning system further optimizes the BP neural network by using an improved PSO algorithm (Improved Particle Swarm Optimization), mainly to improve the weight and bias parameters of the network, so as to achieve better training effect and prediction performance. First, a particle swarm is randomly generated, and each particle represents a set of parameters of the BP neural network, and the parameters are a weight matrix Wzl and bias vector b zl where the position of the particle P zl As shown in equation (22):

[0098] P zl = [W zl1 , W zl2 , …, W zlg , b zl1 , b zl2 , …, b zlj ] (22)

[0099] where g is the number of weight rectangles and j is the number of bias vectors.

[0100] Optionally, the intelligent early warning system trains a BP neural network for each particle zlusing its parameters and calculates the fitness function, as shown in equation (23):

[0101]

[0102] where MSE zl represents the fitness value of the zlthparticle, usually the mean square error, N is the number of training samples, t k is the target output of the kthsample, o zlk is the predicted output of the kthsample using the parameters of the particle zl.

[0103] Optionally, if the fitness of the particle zl is better than the previously recorded individual optimal fitness then its individual optimal position is updated, i.e. if where represents the best fitness value found so far by the zlthparticle, represents the best position found so far by the zlthparticle, and if MSE zl is the smallest among all particles, then the global optimal position is updated, i.e. if then P best = P zl where represents the best fitness value found so far by all particles, P best represents the global optimal position, i.e. the parameter set corresponding to the smallest MSE among all particles.

[0104] Optionally, the intelligent early warning system updates the speed and position of the particle, as shown in equation (24):

[0105]

[0106] where φ is the inertia weight used to balance exploration and exploitation, c1 and c2 represent the cognitive and social acceleration constants, r 1zl and r 2zl are random numbers for particle zl, generated from a uniform distribution [0, 1], v zl (t) represents the velocity vector of the zl-th particle at time t, v zl (t+1) represents the new velocity vector of the zl-th particle at time t+1, i.e., the velocity of the next iteration, P zl (t+1) represents the new position vector of the zl-th particle at time t+1, which is updated according to its new velocity v zl (t+1), P zl (t) represents the position of the zl-th particle in the search space at time t.

[0107] Optionally, the intelligent early warning system sets a maximum number of iterations J CHG , and repeats the updating of the fitness value of the particle and the optimal position of the individual and the global optimal position and the position and velocity of the particle until the maximum number of iterations J CHG is reached, and the updating is stopped.

[0108] From the above, it can be seen that the intelligent early warning system can automatically adjust the learning rate through the initial learning rate and the decay factor, which helps to quickly learn in the early stage of training and gradually reduce the learning rate in the later stage of training to refine the adjustment of the weight, thereby improving the convergence speed of the network. At the same time, the number of hidden layers and the number of neurons in each hidden layer are determined according to the number of neurons in the input layer and the output layer, which helps the network to better capture complex features in the data. The improved PSO algorithm further optimizes the neural network, improves the prediction accuracy and response speed of the model, and makes the early warning more timely. Finally, by using a multi-modal data set for training, the system can process and learn different types and forms of data, which helps to improve the adaptability and accuracy of the model to different data sources. Through multiple iterations of training, the model is optimized, and each iteration is adjusted based on the results of the previous iteration, which can gradually improve the performance of the model until a satisfactory level of safety state prediction is achieved.

[0109] In an optional embodiment, the intelligent early warning system divides the multi-modal data set into N subsets, where N is an integer greater than 1, and performs K rounds of test operations on the safety state neural network to obtain K prediction results, where each round of test operation is used to select any one of the N subsets as a target test set and the subsets other than the target test set as a target training set, and the safety state neural network is tested using the target training set, and the target test set is predicted using the tested safety state neural network to obtain a prediction result, and K accuracies of the safety state neural network are determined according to the K prediction results, where the Jth accuracy in the K accuracies represents the prediction accuracy of the safety state neural network in the Jth round of test operation in the K rounds of test operations, and then the average of the K accuracies is calculated as the prediction accuracy of the safety state neural network.

[0110] Optionally, the intelligent early warning system uses cross-validation to evaluate the performance of the BP neural network model, divides the multi-modal data set D into N subsets, each of which is approximately equal in size, denoted as D1, D2, …, DN. k The safety state neural network is tested for K rounds, and in each cycle, one of the subsets is selected as a test set D tese , and all other subsets are used as training sets for model training. In the first cycle, D1 is used to test the safety state neural network for one round, in the second cycle, D2 is used to test the safety state neural network for two rounds, and so on, to obtain K prediction results. According to the K prediction results and the true labels of each test set, K accuracies of the safety state neural network are determined, as shown in equation (25):

[0111]

[0112] Where Acc represents the accuracy, TP i is the number of true positives in the ith cycle, TN i is the number of true negatives in the ith cycle, FP i is the number of false positives in the ith cycle, and FN i is the number of false negatives in the ith cycle.

[0113] Optionally, the intelligent early warning system calculates the average of the accuracies in all k cycles to obtain the prediction accuracy of the model, as shown in equation (26):

[0114]

[0115] Where AAy is the average of the accuracy Acc.

[0116] Optionally, after each cycle, the test set D teseThe performance of the model is evaluated, and the overall performance of the model is obtained, as shown in formula (27):

[0117] Pfae=f evalue (M ME ,D tese ) (27)

[0118] Wherein, M ME represents the BP neural network model, Pfae is the performance index after evaluation, f evalue is the evaluation function.

[0119] As can be seen from the above, the intelligent early warning system divides the data set into multiple subsets, so that the model can be trained and tested under different data configurations, which helps to evaluate the performance of the model under different conditions, and through the use of cross-validation, the model performance can be evaluated under different conditions to ensure its stability and effectiveness in diversified actual working environment. By calculating the average accuracy of multiple test rounds, the system can obtain a comprehensive prediction accuracy, and also provides a quantitative index to evaluate the performance of the model. By analyzing the accuracy under different test rounds and the overall performance of the model, the system can identify which aspects of the model have better prediction effect and which aspects need to be improved, thereby providing guidance for further optimization of the model.

[0120] In an optional embodiment, the intelligent early warning system calculates the absolute difference between the actual safety score and the target score, and then updates the early warning threshold according to the absolute difference and the adjustment coefficient to obtain the target early warning threshold. Finally, the early warning information is generated according to the target early warning threshold and the actual safety score.

[0121] Optionally, the intelligent early warning system analyzes the output of the BP neural network and automatically adjusts the early warning threshold to adapt to the changing working environment, working personnel state, etc. in the limited space.

[0122] Optionally, the intelligent early warning system calculates the deviation between the model output safety score S zs and the actual safety score S acl to obtain the deviation value, as shown in formula (28):

[0123] e sc =|S zs -S acl | (28)

[0124] Wherein, e sc represents the deviation value.

[0125] Optionally, the intelligent early warning system adjusts the early warning threshold G TC according to the deviation value and uses cumulative error, as shown in formula (29):

[0126] G new = G TC + v ·∑e sc (29)

[0127] wherein G new is the adjusted early warning threshold, v is an adjustment coefficient which determines the speed of threshold adjustment, v is a small positive number, ∑e sc represents the value of the sum of all deviation values e sc .

[0128] From the above, the intelligent early warning system first calculates the absolute difference between the actual safety score and the target score, quantifying the gap between the two, then updates the early warning threshold according to the calculated absolute difference and the preset adjustment coefficient, through the establishment of the adjustment mechanism of the dynamic threshold, makes the early warning system can adapt to the changes of the working environment and the state of the workers, improves the adaptability and accuracy of the early warning, can more accurately reflect the current safety situation, reduces the false alarm and the missed alarm, at the same time improves the adaptability and flexibility of the system.

[0129] In an alternative embodiment, the intelligent early warning system determines a first value according to the environmental data in the limited space and the weight of the environmental data, wherein the first value is used to represent the risk intensity caused by the environmental factors in the limited space, then determines a second value according to the personnel state data of the workers in the limited space and the contribution factor, wherein the contribution factor is used to represent the probability of risk caused by the physical factors of the workers, wherein the second value is used to represent the risk intensity caused by the physical state of the workers, then takes the absolute difference between the first value and the second value as the target difference, and determines the risk type in the limited space according to the target early warning threshold, the first value, the second value and the target difference, wherein the risk type is used to distinguish the source category of the limited space risk, then determines the risk level in the limited space according to the actual safety score and at least one preset risk threshold, wherein the risk level is used to represent the risk degree in the limited space, and finally generates the early warning information according to the risk type and the risk level.

[0130] Optionally, the intelligent early warning system establishes a contribution mechanism to determine the main risk source by comparing the contribution values, first calculates the total contribution degree of environmental factors (the above-mentioned first value), as shown in formula (30):

[0131] E xd = γ 1 · T + γ 2 · H + γ 3 · G + γ 4 · OC (30)

[0132] wherein γ 1 , γ 2 , γ3 and gamma 4 are weights corresponding to temperature T, humidity H, gas concentration G and oxygen content OC respectively.

[0133] Optionally, the intelligent early warning system calculates the total contribution degree of the personal factors of the workers (the second numerical value described above), as shown in formula (31):

[0134]

[0135] wherein, P heth represents the health condition, which is the average value of heart rate HR, blood pressure BP and blood oxygen saturation SP, is the weight (contribution factor described above) of P heth .

[0136] Optionally, the intelligent early warning system calculates the absolute difference between the total contribution degree of the environmental factors and the total contribution degree of the personal factors, and then determines the risk type in the limited space according to the adjusted early warning threshold, the absolute difference, the total contribution degree of the environmental factors, and the total contribution degree of the personal factors, determines the risk level in the limited space according to the actual safety score and the preset risk threshold, and finally generates the early warning information according to the risk type and the risk level.

[0137] As can be seen from the above, the intelligent early warning system can analyze the environmental data and the state data of the workers in the limited space, evaluate the risk intensity caused by the environmental factors and the physical factors of the workers, quantify the risk intensity by determining the first numerical value and the second numerical value, so that the risk evaluation is more accurate and operable, and according to the risk type and the risk level, the system can generate corresponding early warning information to provide decision support for safety management.

[0138] In an optional embodiment, the intelligent early warning system determines that the first type of risk occurs in the limited space when the first numerical value is greater than the second numerical value and the target difference value is greater than the target early warning threshold, wherein the first type of risk is used to represent the risk caused by the environmental factors, determines that the second type of risk occurs in the limited space when the first numerical value is less than the second numerical value and the target difference value is greater than the target early warning threshold, wherein the second type of risk is used to represent the risk caused by the physical factors of the workers, and determines that the third type of risk occurs in the limited space when the target difference value is less than or equal to the target early warning threshold, wherein the third type of risk is used to represent the risk caused by the environmental factors and the physical factors of the workers.

[0139] Optionally, the first type of risk represents the environmental factor risk, such as excessive temperature, excessive humidity, excessive gas concentration and low oxygen content, the second type of risk represents the personal factor risk, such as abnormal heart rate, abnormal blood pressure and abnormal blood oxygen saturation, and the third type of risk is a comprehensive risk, which is a combination of the environmental factor risk and the personal factor risk.

[0140] Optionally, if E xd R xd , and E xd -R xd , G new , it is defined as a class of risk; if R xd >E xd , and R xd -E xd , G new , it is defined as a class of risk; if |E xd -R xd |≤G new , it is defined as a class of risk.

[0141] Optionally, the intelligent early warning system positions the risk level as four risk levels of slight, medium, severe and extreme, and sets the threshold values of the four risk levels of slight, medium, severe and extreme as FB1, FB2, FB3 and FB4, and then obtains the risk instruction through threshold comparison, if S acl <FB1, it is determined as a slight risk, and a slight risk instruction is issued; if FB1≤S acl <FB2, it is determined as a medium risk, and a medium risk instruction is issued; if FB2≤S acl <FB3, it is determined as a severe risk, and a severe risk instruction is issued; if FB3≤S acl <FB4, it is determined as an extreme risk, and an extreme risk instruction is issued.

[0142] Optionally, in the case of a class of risk: the warning light emits purple light at the slight level, reminding the staff to pay attention to environmental changes; the ventilation system is automatically adjusted at the medium level, and the alarm emits an alarm sound; the emergency lighting is started at the severe level, the ventilation system is automatically adjusted to the maximum, and the alarm emits evacuation information; all unnecessary power is automatically turned off at the extreme level, and the emergency evacuation and rescue program is started.

[0143] Optionally, in the case of a class of risk: the warning light emits blue light at the slight level, reminding the staff to pay attention to personal health status; the alarm emits rest information at the medium level, allowing the staff to rest; emergency medical examination is arranged at the severe level, and the alarm emits evacuation information; immediate first aid is provided at the extreme level, and the alarm emits emergency evacuation information.

[0144] Optionally, in the case of a class of risk: the warning light emits red light at the slight level, reminding the staff to pay attention to environmental and personal health status; the ventilation system is automatically adjusted at the medium level, and the alarm emits rest information, allowing the staff to rest; the emergency lighting is started at the severe level, and the alarm emits evacuation information. All unnecessary power is automatically turned off at the extreme level, immediate first aid is provided, and the alarm emits emergency evacuation information.

[0145] Optionally, if multiple risks are detected at the same time, the intelligent early warning system processes the early warning in priority order, and the order of the risk type level from high to low is three-class risk, two-class risk, and one-class risk, and the order of the risk level from high to low is extreme risk level, serious risk level, medium risk level, and slight risk level.

[0146] From the above, it can be seen that when the intelligent early warning system detects potential risks, it not only issues an early warning, but also automatically triggers corresponding emergency response measures according to the risk level and type, realizing seamless connection from early warning to emergency disposal and effectively reducing the risk of accidents.

[0147] According to another aspect of the embodiments of the present application, an embodiment of a safety early warning device for a limited space is also provided. Figure 2 is a schematic diagram of an optional safety early warning device for a limited space according to an embodiment of the present application, as Figure 2 shown, the safety early warning device for a limited space includes a determination unit 201, an input unit 202, and a generation unit 203.

[0148] Optionally, the input unit 202 includes a first acquisition subunit, a second processing subunit, a first determination subunit, and a second determination subunit. The first acquisition subunit is configured to acquire historical data in the limited space, wherein the historical data at least includes historical environmental data and historical personnel state data of workers in the limited space; the second processing subunit is configured to extract key features from the historical data, and fuse the key features according to the weight of each key feature to obtain feature vectors at different times, wherein the key features at least include historical environmental parameters and physiological characteristics of historical workers in the limited space, wherein the historical environmental parameters at least include temperature, humidity, gas concentration, and oxygen content, and the physiological characteristics of the historical workers at least include heart rate, blood pressure, and blood oxygen saturation; the first determination subunit is configured to obtain a multi-modal data set according to the feature vectors and the risk score value of each feature vector; and the second determination subunit is configured to train a safety state neural network based on the multi-modal data set.

[0149] Optionally, the second determining subunit comprises a first obtaining module, a first determining module, a second determining module and a first training module. The first obtaining module is configured to obtain an initial learning rate, an input layer neuron quantity and an output layer neuron quantity of the initial neural network, wherein the initial learning rate is used to represent an initial setting update frequency of a weight and a bias vector in the initial neural network; the first determining module is configured to determine a target learning rate of the initial neural network according to the initial learning rate and a decay factor, wherein the target learning rate is used to improve the convergence speed of the initial neural network, and the decay factor is used to represent a decay rate of the initial learning rate; the second determining module is configured to determine a target quantity of hidden layers and a neuron quantity of each hidden layer of the initial neural network according to the input layer neuron quantity and the output layer neuron quantity, wherein the hidden layers are located between an input layer and an output layer of the initial neural network, and the target quantity and the neuron quantity are used to improve the performance of the initial neural network; and the first training module is configured to perform multiple iteration training on the initial neural network according to the multi-modal data set, and obtain the safety state neural network through training.

[0150] Optionally, the second determining subunit comprises a first processing module, a second processing module, a third determining module and a first calculation module. The first processing module is configured to divide the multi-modal data set into N subsets, wherein N is an integer greater than 1; the second processing module is configured to perform K rounds of test operations on the safety state neural network to obtain K prediction results, wherein each round of test operation is used to select any one of the N subsets as a target test set, and the subsets other than the target test set are used as a target training set, the safety state neural network is tested for one round using the target training set, and the target test set is predicted using the tested safety state neural network to obtain one prediction result; the third determining module is configured to determine K accuracies of the safety state neural network according to the K prediction results, wherein the Jth accuracy in the K accuracies is used to represent the prediction accuracy of the safety state neural network in the Jth test operation in the K test operations; and the first calculation module is configured to calculate an average value of the K accuracies as a prediction accuracy of the safety state neural network.

[0151] Optionally, the generating unit 203 comprises a first calculation subunit, a third processing subunit and a first generating subunit. The first calculation subunit is configured to calculate an absolute difference value of the actual safety score and the target score; the third processing subunit is configured to update the early warning threshold according to the absolute difference value and an adjustment coefficient to obtain a target early warning threshold; and the first generating subunit is configured to generate early warning information according to the target early warning threshold and the actual safety score.

[0152] Optionally, the first generating subunit comprises a fourth determining module, a fifth determining module, a sixth determining module, a seventh determining module, an eighth determining module, and a first generating module. The fourth determining module is configured to determine a first value according to the environmental data in the limited space and the weight of the environmental data, wherein the first value is used to represent the risk intensity caused by the environmental factors in the limited space. The fifth determining module is configured to determine a second value according to the personnel state data of the workers in the limited space and a contribution factor, wherein the contribution factor is used to represent the probability of the risk caused by the physical factors of the workers, and the second value is used to represent the risk intensity caused by the physical state of the workers. The sixth determining module is configured to take the absolute difference between the first value and the second value as a target difference value. The seventh determining module is configured to determine a risk type in the limited space according to the target early warning threshold, the first value, the second value, and the target difference value, wherein the risk type is used to distinguish the source category of the risk in the limited space. The eighth determining module is configured to determine a risk level in the limited space according to the actual safety score and at least one preset risk threshold, wherein the risk level is used to represent the risk degree in the limited space. The first generating module is configured to generate the early warning information according to the risk type and the risk level.

[0153] Optionally, the seventh determining module comprises a first determining submodule, a second determining submodule, and a third determining submodule. The first determining submodule is configured to determine that a first type of risk occurs in the limited space in a case where the first value is greater than the second value and the target difference value is greater than the target early warning threshold, wherein the first type of risk is used to represent the risk caused by the environmental factors. The second determining submodule is configured to determine that a second type of risk occurs in the limited space in a case where the first value is less than the second value and the target difference value is greater than the target early warning threshold, wherein the second type of risk is used to represent the risk caused by the physical factors of the workers. The third determining submodule is configured to determine that a third type of risk occurs in the limited space in a case where the target difference value is less than or equal to the target early warning threshold, wherein the third type of risk is used to represent the risk caused by the environmental factors and the physical factors of the workers.

[0154] According to another aspect of the present application, an electronic device is also provided, wherein the electronic device comprises one or more processors and a memory, and the memory is configured to store one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors are caused to perform the safety early warning method for the limited space.

[0155] According to another aspect of the present application, a computer readable storage medium is also provided, wherein the computer readable storage medium stores a computer program, and when the computer program runs, the device where the computer readable storage medium is located performs the safety early warning method for the limited space.

[0156] The above-mentioned sequence numbers of the embodiments of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0157] In the above-mentioned embodiments of the present application, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0158] In the several embodiments provided by the present application, it should be understood that the disclosed technology can be implemented in other ways. Of course, the unit embodiment described above is only illustrative, and for example, the division of units can be a logical function division, and there can be another division manner in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, unit or module, and can be electrical or other forms.

[0159] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed to multiple units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0160] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of software functional unit.

[0161] The integrated unit, if realized in the form of software functional unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the part of the prior art that makes a contribution or the whole or part of the technical solutions can be embodied in the form of software product, which is stored in a storage medium and includes a plurality of instructions for making a computer device (which can be a personal computer, a server or a network device, etc.) execute all or part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and various program code storage media.

[0162] The above merely describes the preferred embodiments of the present application, and it should be pointed out that, for those skilled in the art, some improvements and refinements can be made without departing from the principles of the present application, and these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. A safety early warning method for confined spaces, characterized in that, include: The actual safety score within the confined space is determined based on the target data within the confined space, wherein the target data includes at least environmental data, equipment data, personnel status data, and operational data within the confined space; The actual safety score is input into the safety state neural network to obtain the target score, wherein the target score is used to characterize the overall safety risk level within the finite space; Warning information is generated based on the warning threshold, the actual safety score, and the target score, wherein the warning threshold is periodically adjusted based on the actual safety score and the target score; The process of generating warning information based on a warning threshold, the actual safety score, and the target score includes: calculating the absolute difference between the actual safety score and the target score; updating the warning threshold based on the absolute difference and an adjustment coefficient to obtain a target warning threshold; and generating warning information based on the target warning threshold and the actual safety score. The method of generating warning information based on the target warning threshold and the actual safety score includes: determining a first value based on environmental data within the confined space and the weight of the environmental data, wherein the first value is used to characterize the risk intensity caused by environmental factors within the confined space; determining a second value based on the personnel status data of the workers within the confined space and a contribution factor, wherein the contribution factor is used to characterize the probability of risk caused by the physical factors of the workers, and the second value is used to characterize the risk intensity caused by the physical condition of the workers; using the absolute difference between the first value and the second value as a target difference; determining the risk type within the confined space based on the target warning threshold, the first value, the second value, and the target difference, wherein the risk type is used to distinguish the source category of the risk in the confined space; determining the risk level within the confined space based on the actual safety score and at least one preset risk threshold, wherein the risk level is used to characterize the degree of risk in the confined space; and generating warning information based on the risk type and the risk level.

2. The safety early warning method for confined spaces according to claim 1, characterized in that, The secure state neural network is trained through the following steps: Acquire historical data within the confined space, wherein the historical data includes at least historical environmental data and historical personnel status data within the confined space; Key features are extracted from the historical data, and the key features are fused according to the weight of each key feature to obtain feature vectors at different times. The key features include at least the historical environmental parameters within the limited space and the physiological characteristics of the historical workers. The historical environmental parameters include at least temperature, humidity, gas concentration, and oxygen content, and the physiological characteristics of the historical workers include at least heart rate, blood pressure, and blood oxygen saturation. A multimodal dataset is obtained based on the feature vectors and the risk score value of each feature vector; The safe state neural network is trained based on the multimodal dataset.

3. The safety early warning method for confined spaces according to claim 2, characterized in that, The safe state neural network is trained based on the multimodal dataset, including: Obtain the initial learning rate, the number of neurons in the input layer, and the number of neurons in the output layer of the initial neural network, wherein the initial learning rate is used to characterize the update frequency of the initial settings of the weights and bias vectors in the initial neural network; The target learning rate of the initial neural network is determined based on the initial learning rate and the decay factor, wherein the target learning rate is used to improve the convergence speed of the initial neural network, and the decay factor is used to characterize the rate at which the initial learning rate decays. The target number of hidden layers and the number of neurons in each hidden layer of the initial neural network are determined based on the number of neurons in the input layer and the number of neurons in the output layer, wherein the hidden layers are located between the input layer and the output layer of the initial neural network, and the target number and the number of neurons are used to improve the performance of the initial neural network; The initial neural network is trained iteratively multiple times based on the multimodal dataset to obtain the safe state neural network.

4. The safety early warning method for confined spaces according to claim 2, characterized in that, After training the safety state neural network based on the multimodal dataset, the finite space safety early warning method further includes: The multimodal dataset is divided into N subsets, where N is an integer greater than 1; The security state neural network is subjected to K rounds of testing to obtain K prediction results. In each round of testing, any subset of the N subsets is selected as the target test set, and the subsets other than the target test set are used as the target training set. The security state neural network is tested once using the target training set, and the tested security state neural network is used to predict the target test set to obtain a prediction result. The K accuracies of the safety state neural network are determined based on the K prediction results, wherein the Jth accuracy of the K accuracies is used to characterize the prediction accuracy of the safety state neural network under the Jth test operation in the K rounds of test operations; The average of the K accuracies is calculated as the prediction accuracy of the safety state neural network.

5. The safety early warning method for confined spaces according to claim 1, characterized in that, Determining the risk type within the finite space based on the target warning threshold, the first value, the second value, and the target difference includes: If the first value is greater than the second value and the target difference is greater than the target warning threshold, it is determined that a first type of risk has occurred in the confined space, wherein the first type of risk is used to characterize the risk caused by environmental factors; If the first value is less than the second value and the target difference is greater than the target warning threshold, it is determined that a second type of risk has occurred in the confined space, wherein the second type of risk is used to characterize the risk caused by the physical factors of the operator; If the target difference is less than or equal to the target warning threshold, a third type of risk is determined to exist within the confined space, wherein the third type of risk is used to characterize the risk caused by the combined environmental factors and the physical factors of the workers.

6. A confined space safety early warning device, used to execute the confined space safety early warning method according to any one of claims 1 to 5, characterized in that, include: The determining unit determines the actual safety score within the confined space based on target data within the confined space, wherein the target data includes at least environmental data, equipment data, personnel status data, and work data within the confined space; The input unit inputs the actual safety score into the safety state neural network to obtain the target score, wherein the target score is used to characterize the overall safety risk level within the finite space; The generation unit generates warning information based on the warning threshold, the actual safety score, and the target score, wherein the warning threshold is periodically adjusted based on the actual safety score and the target score.

7. An electronic device, characterized in that, It includes one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to perform the confined space security warning method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the computer program is executed, the device containing the computer-readable storage medium performs the confined space security warning method according to any one of claims 1 to 5.

Images