Intelligent prediction and early warning method and device for safety risk of underground personnel and electronic equipment

Abstract

The invention provides an intelligent prediction and early warning method and device for safety risks of underground personnel and electronic equipment. The method comprises the following steps: acquiring vital sign information and working environment information of underground working personnel within first time; according to the vital sign information and the working environment information, a pre-trained vital sign change trend prediction model is adopted to predict the vital sign change trend of the working personnel, and the vital sign change trend of the working personnel is obtained; according to the vital sign change trend and the working environment information, vital sign comprehensive evaluation information of the working personnel is determined; an emergency is detected according to the vital sign information, and a detection result for the emergency is obtained; and according to the vital sign comprehensive evaluation information of the operator, the vital sign change trend and the detection result for the emergency, performing graded early warning to obtain the safety risk early warning level of the operator. By implementing the embodiment of the invention, the safety of coal mine operation can be improved.

Description

Technical Field

[0001] The present disclosure relates to the field of coal mine safety and the field of artificial intelligence technologies such as deep learning, and particularly relates to an intelligent prediction and early warning method, device, electronic device, and storage medium for underground personnel safety risks. Background Art

[0002] The operation environment in coal mines is complex, and disaster accidents occur frequently, such as gas explosions, coal dust explosions, gas outbursts, etc. These disasters not only pose a threat to mine facilities, but also directly threaten the lives of operating personnel. The mine environment usually has characteristics such as high temperature, high humidity, low oxygen, and gas pollution. Under sudden disaster conditions, environmental factors such as gas concentration, temperature, and humidity in the mine change sharply, which may cause sudden changes in the physical health of miners (such as suffocation, heart attacks, etc.). Existing coal mine safety early warning systems mainly rely on monitoring data of the mine environment, such as gas concentration, oxygen content, etc. However, there is insufficient monitoring of miners' vital signs, and there is a lack of an intelligent early warning method that comprehensively considers the mine disaster environment and changes in miners' vital signs. Summary of the Invention

[0003] Embodiments of the present disclosure provide an intelligent prediction and early warning method, device, electronic device, and storage medium for underground personnel safety risks, which can solve the problem of insufficient intelligence in predicting and early warning the life signs of personnel in complex coal mine underground environments with a variety of work types.

[0004] According to a first aspect of the embodiments of the present disclosure, an intelligent prediction and early warning method for underground personnel safety risks is provided, including:

[0005] Obtain the vital sign information and operation environment information of underground operating personnel within a first period of time;

[0006] According to the vital sign information and the operation environment information, use a pre-trained vital sign change trend prediction model to predict the vital sign change trend of the operating personnel, and obtain the vital sign change trend of the operating personnel;

[0007] According to the vital sign change trend and the operation environment information, determine the comprehensive vital sign assessment information of the operating personnel;

[0008] Detect emergencies according to the vital sign information, and obtain a detection result for the emergencies;

[0009] According to the comprehensive vital sign assessment information of the operating personnel, the vital sign change trend, and the detection result for the emergencies, perform hierarchical early warning to obtain the safety risk early warning level of the operating personnel.

[0010] According to a second aspect of the embodiments of the present disclosure, there is provided an intelligent prediction and early warning device for underground personnel safety risks, including:

[0011] An acquisition module, configured to acquire the vital sign information and working environment information of underground workers within a first period of time;

[0012] A prediction module, configured to predict the changing trend of the vital signs of the worker according to the vital sign information and the working environment information, by using a pre-trained prediction model for the changing trend of vital signs, to obtain the changing trend of the vital signs of the worker;

[0013] A determination module, configured to determine the comprehensive evaluation information of the vital signs of the worker according to the changing trend of the vital signs and the working environment information;

[0014] A detection module, configured to detect emergencies according to the vital sign information, to obtain a detection result for the emergencies;

[0015] An early warning module, configured to perform hierarchical early warning according to the comprehensive evaluation information of the vital signs of the worker, the changing trend of the vital signs, and the detection result for the emergencies, to obtain the safety risk early warning level of the worker.

[0016] According to a third aspect of the embodiments of the present disclosure, there is provided an electronic device, including: one or more processors; wherein, the electronic device is configured to execute the intelligent prediction and early warning method for underground personnel safety risks described in the foregoing first aspect.

[0017] According to a fourth aspect of the embodiments of the present disclosure, there is provided a storage medium, which stores instructions that, when running on an electronic device, cause the electronic device to execute the intelligent prediction and early warning method for underground personnel safety risks described in the foregoing first aspect.

[0018] According to a fifth aspect of the embodiments of the present disclosure, there is provided a computer program product, including a computer program that, when executed by an electronic device, implements the steps of the method described in the foregoing first aspect.

[0019] According to the technical solution of the present disclosure, by integrating mine environment data and vital sign data of workers for intelligent prediction and early warning of safety risks, it is possible to solve the technical requirements for predicting and early warning the future vital sign status of personnel under the influence of environmental factors in coal mines, and it is also possible to solve the problem of intelligent construction of coal mines, thereby making up for the deficiency in intelligent prediction and early warning only for the vital sign data of underground personnel, and being used for real-time prediction and early warning of the life safety risks of personnel in complex mine environments, which can improve the safety of coal mine operations.

[0020] Additional aspects and advantages of the present disclosure will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present disclosure. Description of the Drawings

[0021] The above-mentioned and / or additional aspects and advantages of the present disclosure will become apparent and be readily understood from the following description of embodiments in conjunction with the drawings, where:

[0022] Figure 1 is a schematic flowchart of the intelligent prediction and early warning method for underground personnel safety risks provided by an embodiment of the present disclosure;

[0023] Figure 2 is a schematic flowchart of the intelligent prediction and early warning method for underground personnel safety risks provided by an embodiment of the present disclosure;

[0024] Figure 3 is a schematic diagram of the basic structure of the LSTM model provided by an embodiment of the present disclosure;

[0025] Figure 4 is a schematic flowchart of the intelligent prediction and early warning device for underground personnel safety risks provided by an embodiment of the present disclosure;

[0026] Figure 5 is a block diagram of an electronic device according to an embodiment of the present disclosure. Detailed Embodiments

[0027] Embodiments of the present disclosure will be described in detail below. Examples of the embodiments are shown in the drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the drawings are exemplary and are intended to explain the present disclosure and should not be construed as limiting the present disclosure.

[0028] The following describes exemplary embodiments of the present disclosure, including various details of the embodiments of the present disclosure to facilitate understanding. It should be considered that they are merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0029] The terms used in one or more embodiments of the present disclosure are for the purpose of describing specific embodiments only and are not intended to limit one or more embodiments of the present disclosure. The singular forms "a", "the", and "said" used in one or more embodiments of the present disclosure and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in one or more embodiments of the present disclosure refers to and includes any or all possible combinations of one or more of the associated listed items.

[0030] In the embodiments of the present disclosure, "a plurality of" means two or more. In some embodiments, notations such as "at least one of A and B", "A and / or B", "in one case A, in another case B", "in response to one case A, in response to another case B", etc. may, according to the situation, include the following technical solutions: In some embodiments, A (performing A independently of B); in some embodiments, B (performing B independently of A); in some embodiments, selecting to perform from A and B (A and B are selectively performed); in some embodiments, A and B (both A and B are performed). The same is true when there are more branches such as A, B, C, etc.

[0031] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data for analysis, stored data, displayed data, etc.), and signals involved in the present disclosure are all authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions.

[0032] It is worth noting that in the embodiments of the present disclosure, certain industry-existing solutions such as certain software, components, models, etc. may be mentioned. They should be considered exemplary, and their purpose is only to illustrate the feasibility in the implementation of the technical solutions of the present disclosure, but it does not mean that the applicant has already or necessarily used this solution.

[0033] The intelligent prediction and early warning method, device, and electronic device for underground personnel safety risks in the embodiments of the present disclosure will be described below with reference to the accompanying drawings.

[0034] Among them, it should be noted that the execution subject of the intelligent prediction and early warning method for underground personnel safety risks in the embodiments of the present disclosure can be an intelligent prediction and early warning device for underground personnel safety risks. This device can be implemented in a software and / or hardware manner and can be configured in an electronic device. Exemplarily, the electronic device may include but is not limited to a terminal, a server, etc.

[0035] Figure 1 It is a schematic flowchart of the intelligent prediction and early warning method for underground personnel safety risks provided for the embodiments of the present disclosure. As Figure 1As shown, the intelligent prediction and early warning method for underground personnel safety risks may include but is not limited to the following steps.

[0036] In step 101, obtain the vital sign information and working environment information of underground workers within the first period of time.

[0037] Optionally, the vital sign information and working environment information of coal mine underground workers within the first period of time (such as a period of time, such as 10 minutes, etc.) can be obtained. Exemplarily, environmental sensors are set in the coal mine underground, such as temperature and humidity sensors, oxygen sensors, harmful gas sensors, illumination intensity sensors, noise sensors, etc., and the working environment information where the underground workers are located is collected through the environmental sensors. Underground workers can wear wearable devices (such as smart bracelets or smart clothing, etc.), and the vital sign information of the underground workers is collected through the wearable devices. Exemplarily, the environmental sensors set in the coal mine underground and the wearable devices worn by the underground workers can be communicatively connected to an electronic device, and the electronic device can receive the working environment information collected by the environmental sensors through this communication connection and receive the vital sign information collected by the wearable devices worn by the underground workers. It can be understood that the above-obtained vital sign information and working environment information can both be time series data.

[0038] In some embodiments, when obtaining the working environment data collected by the environmental sensors and the vital sign data collected by the wearable devices, the obtained data can be preprocessed to obtain the preprocessed vital sign information and working environment information. Exemplarily, the wavelet denoising method can be used to remove the noise in the data collected by the environmental sensors and the wearable devices respectively; the interpolation method or other methods can be used to fill in the missing data to ensure data integrity; all the data collected by the environmental sensors and the wearable devices respectively are aligned to the same timestamp to ensure the consistency of the time series; the data is normalized or standardized to eliminate the influence brought by different dimensions.

[0039] In some embodiments, the above vital sign information may include but is not limited to body temperature, heart rate, blood pressure, respiratory rate, blood oxygen saturation, etc. The above working environment information may include but is not limited to temperature, humidity, oxygen concentration, harmful gas concentration, noise, illumination intensity, wind speed, etc.

[0040] In step 102, according to the vital sign information and the working environment information, use a pre-trained vital sign change trend prediction model to predict the vital sign change trend of the worker, and obtain the vital sign change trend of the worker.

[0041] In some embodiments, data fusion and deep feature extraction may be performed based on the vital sign information and the working environment information to obtain a deep feature representation after the fusion of the vital sign information and the working environment information; the deep feature representation is input into a pre-trained vital sign change trend prediction model to predict the vital sign change trend of the operator, and the vital sign change trend of the operator is obtained. Among them, the structure of the vital sign change trend prediction model may include LSTM (Long short-term memory), that is: the vital sign change trend prediction model may be constructed based on LSTM. LSTM is a neural network architecture specifically used to process time series data, and it can capture long-term dependencies in time series data. The vital sign change trend prediction model may be trained based on training data, and the training data may include the vital sign information of coal miners at different times and the working environment information at the same time, as well as the clinical medical data of the affected workers. The input of the vital sign change trend prediction model is the above-mentioned deep feature representation, and the output of the vital sign change trend prediction model is the vital sign change trend.

[0042] In step 103, based on the vital sign change trend and the working environment information, the comprehensive vital sign assessment information of the operator is determined.

[0043] In some embodiments, the maximum survival period of the human body corresponding to the vital sign change trend may be determined based on the quantitative relationship between the preset vital sign index and the maximum survival period of the human body; the environmental risk score information is obtained by weighted summation according to the working environment information and its weight; according to the maximum survival period of the human body and the environmental risk score information, the comprehensive vital sign assessment information of the operator is obtained by using a preset comprehensive vital sign assessment function.

[0044] In some embodiments, the quantitative relationship between the above-mentioned vital sign indicators and the maximum human survival period can be established based on the multiple regression analysis method, with the vital sign indicators as independent variables and the maximum human survival period as the dependent variable. Exemplarily, the quantitative relationship between the vital sign indicators and the maximum human survival period can be pre-established based on historical sample data using the multiple regression analysis method. In one possible implementation, the vital sign information of coal miners at different times, the working environment information at the same time, and the clinical medical data of the affected workers can be obtained. The clinical medical data is mainly the human life survival time. Based on the working environment data, human vital sign data, and clinical medical data, analyze the quantitative relationships between respiratory entropy, anaerobic threshold, oxygen pulse, etc. and the maximum human survival period, denoise and standardize the data, and use the association rule mining algorithm to optimize the vital sign indicators and environmental indicators based on medical theory. Input vital sign indicators such as respiratory entropy, anaerobic threshold, and oxygen pulse into the multiple regression analysis algorithm to obtain the calculation method of the maximum human survival period. This calculation method of the maximum human survival period is the quantitative relationship between the vital sign indicators and the maximum human survival period. Assuming that the quantitative relationship is a linear regression, with N samples (such as worker samples) and M vital sign indicator independent variables, this quantitative relationship can be expressed as:

[0045] Y n =f(X NM )=f(X n1 ,X n2 ,…,X nM )=β1X n1 +β2X n2 +…+β M X nM +β0 (1)

[0046] Where Y n is the maximum human survival period (in hours); X nm ∈X NM is the value of the m-th vital sign indicator of the n-th sample, n = 1, 2, …, N, m = 1, 2, …, M; β0 is the intercept, and β1, β2, …, β M are the regression coefficients, representing the influence of each vital sign indicator independent variable on the dependent variable (the maximum human survival period), that is, the weights of each vital sign indicator, and this weight can be determined based on historical sample data and the multiple regression analysis method. Exemplarily, in actual monitoring applications, when obtaining the vital sign information of underground workers, the value of the corresponding vital sign indicator can be determined using the predicted trend of vital sign changes, and substituting the value of this vital sign indicator into the above formula (1) can obtain the maximum human survival period corresponding to the predicted trend of vital sign changes.

[0047] In some embodiments, the weight of the above-mentioned operation environment information can be pre-determined. For example, it can be determined based on the entropy weight method. That is to say, the entropy weight method can be used to obtain the weight of the operation environment information. For example, the entropy weight method is used for weighting to obtain the environmental risk score information. Exemplarily, assuming there are N samples and K environmental indicators, the implementation process of obtaining the weight of the operation environment information by the entropy weight method can be as follows:

[0048]

[0049] Among them, x ij is the original value of the j-th environmental indicator of the i-th sample, min i x ij is the minimum of the j-th environmental indicator, max i x ij is the maximum value of the j-th environmental indicator. The symbol represents traversing all rows (i rows of data) under the j-th column, that is, the j-th environmental indicator, and selecting the minimum and maximum values inside; z ij is the standardized value of the j-th environmental indicator of the i-th sample; p ij is the proportional value of the j-th environmental indicator of the i-th sample to the j-th indicator value of all samples, is the sum of the standardized values of the j-th environmental indicator of all samples, e j is the information entropy value of the j-th environmental indicator, and the general constant is the sum of the product of the proportional value of the j-th environmental indicator value of all samples and its natural logarithm; α j is the weight of the j-th environmental indicator. Optionally, the environmental risk score information can be calculated by the following formula (6):

[0050]

[0051] Among them, S i is the comprehensive score of the i-th sample, that is, the environmental risk score information. Exemplarily, in actual monitoring applications, when obtaining the operation environment information of underground workers, the numerical value of the corresponding environmental indicator can be determined based on the operation environment information, and the environmental risk score information can be calculated based on the numerical value of the environmental indicator using the above formulas (2) to (6).

[0052] In the embodiments of the present disclosure, when obtaining the maximum survival period of the human body and the environmental risk score information corresponding to the predicted change trend of the vital signs, the comprehensive vital sign evaluation information of the worker can be calculated using the vital sign comprehensive evaluation function according to the maximum survival period of the human body and the environmental risk score information. Exemplarily, the vital sign comprehensive evaluation function can be expressed as follows:

[0053] Zn = β·Y n + γ·S n (7)

[0054] Wherein, Z n is the comprehensive assessment information of the vital signs of the nth sample, such as the comprehensive assessment score of vital signs. β and γ are the respective weights of the maximum human survival period and the environmental risk score information obtained by the entropy weight method. Exemplarily, they can be empirical values, but are not limited thereto. Or exemplarily, the weight β of the maximum human survival period may be related to the weights β1, β2, …, β M of the vital sign indicators in the above formula (1); the weight γ of the environmental risk score information may be related to the weights α1, α2, …, α K of the environmental indicators in the above formula (6), but is not limited thereto. For example, the weight parameters β and γ can be optimized and adjusted based on the prediction results of the vital sign change trend prediction model. For example, the model parameters in the vital sign change trend prediction model and the weight parameters β and γ are jointly trained. When the vital sign change trend prediction model is completed, the weight parameters β and γ in the formula (7) are also fixed, so that in practical applications, the predicted vital sign change trend, the working environment information (i.e., environmental factor data), and their respective weights can be directly used in the vital sign comprehensive assessment function to calculate the comprehensive assessment information of the vital signs of the operator.

[0055] In step 104, the emergency event is detected according to the vital sign information, and the detection result for the emergency event is obtained.

[0056] It should be noted that sudden events such as falls and entrapment usually have strong suddenness and require a response within a short time. To detect these sudden situations, in some alternative embodiments, a real-time monitoring mechanism based on a threshold can be designed. Exemplarily, a sudden change detection method for vital sign changes can be used, such as detecting the occurrence of an event by monitoring sudden changes in vital sign information such as heart rate, body temperature, and respiratory rate. For example, detecting whether the heart rate has increased sharply or whether the body temperature has exceeded the normal range.

[0057] Assume that the heart rate change rate ΔHR t = |HR t - HR t-1 | is the sudden change index of the heart rate. If the change exceeds a certain threshold θ HR , it is determined as a possible fall or other emergency event. The formula is as follows:

[0058] Optionally, a multi-variable mutation detection method is adopted to detect whether an emergency occurs. That is to say, instead of relying solely on a single vital sign, multiple physiological signals can be combined to design multi-dimensional determination conditions. For example, if multiple vital sign information (such as heart rate, body temperature, respiratory rate, etc.) all mutate, it may mean a more serious emergency (such as poisoning, heart attack, etc.). For example, the formula is expressed as follows: That is to say, the vital sign information can be compared with its corresponding threshold value. If the value of the vital sign information exceeds the corresponding threshold value, it can be determined that the corresponding emergency has occurred, that is, the detection result is that an emergency exists; if the values of all vital sign information do not exceed the corresponding threshold values, it can be determined that no emergency has occurred, that is, the detection result is no emergency. For example, taking the body temperature and heart rate change rate as examples of vital sign information, the mutation index of body temperature can be expressed as follows ΔTemp t =|Tempt t -Tempt t-1 |, θ Temp is the threshold value of body temperature mutation. If the change value of body temperature is greater than the threshold value θ Temp , and the heart rate change rate is greater than the threshold value θ HR , it can be explained that abnormal mutations of body temperature and heart rate have occurred between two time steps.

[0059] In step 105, according to the comprehensive evaluation information of the operator's vital signs, the vital sign change trend, and the detection result for emergencies, a hierarchical early warning is performed to obtain the safety risk early warning level of the operator.

[0060] In some embodiments, according to medical knowledge and the vital sign change trend, the predicted vital sign state can be classified to obtain the vital sign state category, and based on the discrimination conditions associated with a set of multiple early warning levels, the comprehensive evaluation information of the operator's vital signs, the vital sign state category, and the detection result for emergencies, a hierarchical early warning is performed.

[0061] In a possible implementation manner, the predicted vital sign state can be classified based on a set of preset state evaluation functions and the vital sign change trend to obtain the vital sign state category; the vital sign state category is used to determine whether the operator is in a dangerous state. Among them, each state evaluation function represents a mapping relationship, and this mapping relationship can be the mapping relationship between the vital sign change trend and the vital sign state category.

[0062] Exemplarily, the changes in vital signs can be classified based on medical knowledge, such as determining whether the person is in a state of fatigue, heatstroke, abnormal breathing, etc. A state evaluation function S can be used to classify the current vital sign state based on the predicted trend of changes in vital signs. For example, the state evaluation function S can be expressed as follows:

[0063]

[0064] where f1, f2, …, f q+1 are functions that model the trend of changes in vital signs corresponding to each state category based on medical knowledge. Threshold 1, Threshold 2, …, Threshold q+1 are usually thresholds determined through historical data and medical expert knowledge. is the predicted trend of changes in vital signs. For example, in formula (8), "fatigue, " is a state evaluation function, representing the mapping relationship between the trend of changes in vital signs and the vital sign state category. Similarly, "poisoning, are state evaluation functions respectively; is the vital sign state category corresponding to the trend of changes in vital signs. Exemplarily, when the predicted trend of changes in vital signs is obtained, the current vital sign state can be classified through the above formula (8) to obtain the vital sign state category.

[0065] In another possible implementation, based on the trend of changes in vital signs, a classification algorithm can be used to classify the predicted vital sign state to obtain the vital sign state category. Among them, the classification algorithm can be, for example, SVM (Support Vector Machine), decision tree, random forest, etc. Exemplarily, the classification algorithm can be a classifier (or classification model), and the predicted trend of changes in vital signs can be used as the input feature of the classifier (or classification model), and the vital sign state category can be obtained through the classifier (or classification model).

[0066] In the embodiments of the present disclosure, after obtaining the vital sign state category, hierarchical warning can be performed based on the discrimination conditions associated with a set of multiple warning levels, the comprehensive evaluation information of the operator's vital signs, the vital sign state category, and the detection results for emergencies, to obtain the safety risk warning level of the operator.

[0067] Exemplarily, based on the comprehensive evaluation information of vital signs, the vital sign state category, and the detection results for emergencies, the warning level and warning trigger mechanism can be preset. The warning level can be divided into multiple levels. For example, it can be divided into 3 levels, as shown in the following example:

[0068] Warning level 1 (low), the associated discrimination conditions are as follows:

[0069] No emergency, Z m <Preset value 1

[0070] That is, it indicates a slight change in vital signs and no emergency treatment is required.

[0071] Early warning level 2 (medium), and the associated discrimination conditions are as follows:

[0072] No emergency, Preset value 1 < Z n <Preset value 2

[0073] That is, it indicates that the change in vital signs is moderately abnormal and needs to be closely monitored.

[0074] Early warning level 3 (high), and the associated discrimination conditions are as follows:

[0075]

[0076] That is, it indicates that the change in vital signs is severely abnormal, and immediate intervention measures should be taken.

[0077] Optionally, in some embodiments, when the safety risk early warning level meets the early warning trigger mechanism, an early warning signal can be triggered; and / or, according to this safety risk early warning level, intelligent decision-making support suggestions can be provided. Exemplarily, the above-mentioned safety risk early warning level meeting the early warning trigger mechanism can be understood as: this safety risk early warning level reaches the level for triggering an early warning. For example, this early warning trigger mechanism can be: an early warning is triggered when the safety risk early warning level reaches early warning level 2. For example, when hierarchical early warning is performed, when it is determined that the current safety risk early warning level is early warning level 1, no early warning signal is triggered. Another example is that when it is determined that the current safety risk early warning level is early warning level 2 or early warning level 3 (or a higher level), an early warning signal is triggered. For example, it can be sent through text messages, APP push, sound and light alarms, etc. to remind operators and managers to take corresponding emergency measures. When an early warning is triggered, corresponding intelligent decision-making support suggestions, such as ventilation, evacuation, oxygen supplementation, etc., can be provided according to this safety risk early warning level.

[0078] In the above embodiments, by integrating mine environment data and operators' vital sign data for intelligent prediction and early warning of safety risks, the technical requirements for predicting and early warning the future vital sign status of personnel under the influence of environmental factors in coal mines can be solved, and the problem of intelligent construction of coal mines can also be solved. Furthermore, it can make up for the deficiency in the aspect of intelligent prediction and early warning only for underground personnel's vital sign data, be used for real-time prediction and early warning of personnel life safety risks in complex mine environments, and can improve the safety of coal mine operations.

[0079] Figure 2Schematic flow chart of the intelligent prediction and early warning method for underground personnel safety risks provided by the embodiments of the present disclosure. As Figure 2 shown, on the basis of what is shown in Figure 1 , the above-mentioned method for predicting the change trend of the vital signs of operating personnel according to the vital sign information and the working environment information by using a pre-trained prediction model for the change trend of vital signs may include, but is not limited to, the following steps.

[0080] In step 201, based on the vital sign information and the working environment information, a time series with a preset duration is constructed by using the sliding window technique, and the time series may include the vital sign information and the working environment information within a plurality of time windows.

[0081] It can be understood that the basic idea of the sliding window technique is to divide the time series data into windows with a fixed length, and each time window may contain historical data within a period of time. The sliding window obtains data at different time steps by moving the window, and forms a new time series.

[0082] In some embodiments, after obtaining the vital sign information and the working environment information of the underground operating personnel within the first period of time, the sliding window may be used to slide the vital sign information and the working environment information to construct a time series with a preset duration, and the time series may include the vital sign information and the working environment information within a plurality of time windows.

[0083] Exemplarily, the working environment information (or environmental factor data): represents various factors in the working environment (such as temperature, humidity, air quality, etc.), denoted as E t , where t is the time step.

[0084] E t = [E t,1 , E t,2 , …, E t,k (9)

[0085] The vital sign information (or vital sign data): represents data related to the individual's vital signs (such as heart rate, blood pressure, body temperature, etc.), denoted as L t , where t is the time step.

[0086] L t = [L t,1 , L t,2 , …, L t,M (10)

[0087] Assume that the size of the sliding window is w, that is, each window contains data of w time steps. The sliding window technique will sequentially select data with time periods of t = 1, 2, …, T for partitioning.

[0088] For each time step t, a data subset generated by the sliding window contains all data from t - w + 1 to t. Specifically, the time series data of each window includes environmental factor data and vital sign data.

[0089] For time step t, the window data of environmental factors can be expressed as:

[0090] E t-w+1:t =[E t-w+1 ,E t-w+2 ,…,E t (11)

[0091] where E t-w+1:t represents the sequence of environmental factor data from t - w + 1 to t.

[0092] L t-w+1:t =[L t-w+1 ,L t-w+2 ,…,L t (12)

[0093] where L t-w+1:t represents the sequence of vital sign data from t - w + 1 to t.

[0094] The environmental factor data and vital sign data are fused within each time window. For each time step t, the fused data can be expressed as:

[0095] S t =[E t-w+1:t ,L t-w+1:t (13)

[0096] where S t is the fused data at time step t, which contains the environmental factor data and vital sign data from t - w + 1 to t.

[0097] The construction of the entire time series S can be completed by iterating the sliding window at different time steps. The specific steps are as follows:

[0098] The sliding window starts at t = w: that is, starting from the w-th time step, the window can completely contain the data of the previous w time steps. The sliding window ends at t = T: the final position of the sliding window is at time step T. Therefore, the time series S after the construction of the sliding window is expressed as:

[0099] S={S w ,S w+1 ,…,S T}=

[0100] {[E w-w+1:w ,Lw-w+1:w ,[E w-w+2:w+1 ,L w-w+2:w+1 ,…,[E T-w+1:T ,L T-w+1:T (14)

[0101] Among them, each S t , where \(t = w,\cdots,T\) are all windows that integrate environmental factor data and vital sign data.

[0102] The size of the window is \(w\), and as the time step progresses, the sliding window gradually obtains the environmental factor data and vital sign data at each time step.

[0103] Through the above steps, the finally generated time series \(S\) includes the fusion data within multiple time windows. The dimensions of the environmental factor data and vital sign data included in each time window are \((w, K + M)\), where \(K\) is the dimension of the environmental factor data, \(M\) is the dimension of the vital sign data, and \(w\) is the length of the sliding window.

[0104] Therefore, the dimension of the time series \(S\) is \((T - w + 1)\times w\times(K + M)\), which represents the environmental factor data and vital sign data fused within the sliding window at each time step.

[0105] In step 202, based on a pre-trained DBN (Deep Belief Network) model, deep feature extraction of the time series is performed to obtain a deep feature representation.

[0106] Exemplarily, after obtaining the time series \(S\) through the above step 201, the time series \(S\) can be input into a pre-trained DBN model for deep feature extraction of the data to obtain a deep feature representation.

[0107] In some embodiments, the DBN model can be composed of multiple RBMs (Restricted Boltzmann Machines). The number of layer nodes in the multi-layer RBM structure in the DBN model is optimized based on the PSO (Particle Swarm Optimization) algorithm, and each layer of the RBM in the DBN model is unsupervised trained based on the optimized number of nodes. Exemplarily, each layer of the RBM in the DBN model can be regarded as an unsupervised learning model for automatically extracting data features. The RBM is a bipartite graph (hidden layer and visible layer) containing two levels, and there are no internal connections between the nodes in each layer. The goal of the RBM is to learn the potential feature representation of the input data through unsupervised learning. Exemplarily, the energy function \(E\) of the RBM can be:

[0108] E(v,h)=-∑ ib i v i -∑ j c j h j -∑ i,j W ij v i h j (15)

[0109] Among them, v i is the state vector of the i-th visible node, h j is the state vector of the j-th hidden node, and W i,j is the weight between the i-th visible node and the j-th hidden node. b i , c j are the biases of the visible layer and the hidden layer, respectively.

[0110] The goal of the RBM is to minimize the energy function, and the Contrastive Divergence (CD) method is usually used for training.

[0111] In the embodiments of the present disclosure, the PSO algorithm can be used to optimize the number of layer nodes in the RBM. PSO is an optimization algorithm based on swarm intelligence, which searches for the optimal solution by simulating the foraging process of a bird flock. Each "particle" represents a solution, and each particle has a position and a velocity. The implementation process can be as follows:

[0112] Initialize the particle swarm: Assume there are N particles, and each particle represents a solution (i.e., the number of nodes in each layer of the RBM). The solution of each particle is a vector containing multiple values, representing the number of nodes in each layer of the RBM.

[0113] The particle position is represented as follows:

[0114] x i = [x i,1 , x i,2 , …, x i,L (i = 1, 2, …, N) (16)

[0115] Among them, L is the number of layers of the RBM, and x i,j represents the number of nodes of the i-th particle in the j-th layer.

[0116] The velocity of the particle: Each particle has a velocity vector v i , which is used to control the position update of the particle:

[0117] v i = [v i,1 , v i,2 , …, v i,L (i = 1, 2, …, N) (17)

[0118] Objective function: The objective function (i.e., fitness function) is usually based on the reconstruction error of RBM or the training error of DBN. Assume that by training the DBN, the loss value L(x i ) of each particle is obtained, and the goal is to minimize this loss value L(x i ). This loss value L(x i ) is expressed as follows.

[0119] L(x i ) = Loss(x i ) (18)

[0120] The velocity and position of each particle are updated in each iteration:

[0121] Update the particle velocity:

[0122] v i,j (t + 1) = wv i,j (t) + c1r1(p i,j - x i,j (t)) + c2r2(g j - x i,j (t)) (19)

[0123] where w is the inertia weight, which controls the movement of the particle along the current direction; c1 and c2 are learning factors, which control the particle to adjust the direction according to its own experience and the group experience; r1 and r2 are random numbers, ranging from [0, 1]; p i,j is the historical optimal position of particle i in the j-th dimension; g j is the global optimal position in the j-th dimension of the group.

[0124] Update the particle position:

[0125] x i,j (t + 1) = x i,j (t) + v i,j (t + 1) (20)

[0126] Update the global optimal solution: In each iteration, check the objective function value of each particle and update the global optimal solution (the node number configuration corresponding to the minimum loss value):

[0127]

[0128] Construct a multi - layer DBN model using the node number configuration of each layer of RBM optimized by PSO. For each layer of RBM, perform unsupervised training using the optimized number of nodes. The training process usually uses the contrastive divergence (CD) algorithm to train each layer of RBM by maximizing the log - likelihood estimate. For the energy function E(v, h) of each RBM layer, the optimization objective is to minimize the reconstruction error, calculate the gradient using the CD method, and update the weights. After training the RBM layers, use the output of each layer of RBM as the input of the next layer and continue training until the entire DBN training process is completed.

[0129] Assume that the input of the k - th layer is the output h of the previous layer k-1 , then the hidden layer representation h k of the k - th layer can be expressed as:

[0130] h k = f k (W k h k-1 + b k ) (22)

[0131] where W k is the weight matrix of the k - th layer; b k is the bias term of the k - th layer; f k is the activation function (such as sigmoid or ReLU).

[0132] After training all RBM layers, the DBN will provide a deep - feature representation h L (t) (i.e., the output of the last layer of the DBN) for each time step t. Thus, the deep - feature extraction of time - series data can be performed through the pre - trained DBN to obtain the deep - feature representation.

[0133] In step 203, predict the vital - sign change trend of the operator according to the deep - feature representation and the pre - trained vital - sign change - trend prediction model to obtain the vital - sign change trend of the operator.

[0134] In some embodiments, the structure of the vital - sign change - trend prediction model can include LSTM. The deep - feature representation h L (t) extracted by the DBN model through unsupervised learning can be used as the input of the LSTM, and the LSTM can learn the dynamic changes of the time series based on these features.

[0135] Assume that the output feature of the DBN model is an N×D L matrix, where N is the number of samples (time steps), and D L is the feature dimension of the last layer of the DBN model. Represent the output feature of the DBN model as hL (t) is used as the input of the LSTM in the order of the time series. Assume that the input at each time step is The sequence shape of the LSTM input is H L = [h L (1), h L (2), …, h L (T)].

[0136] The goal of the LSTM model is to use the input DBN features to perform a regression task and generate predicted values. Assume that a multi-dimensional continuous target variable y t = [y t,1 , y t,2 , …, y t,q is to be predicted, where y t,j is the j-th target variable at time step t, and q is the number of target variables (such as body temperature, temperature, blood pressure).

[0137] As Figure 3 shown, the basic structure of this LSTM can include an input gate i t , a forget gate f t and an output gate o t , and these gates control the flow of information. For multivariate prediction, the LSTM captures the long-term dependencies of the input sequence through its memory cell C t . Exemplarily, the core calculation formulas of the LSTM are as follows:

[0138] (1) Forget gate: Determines how much past state information to forget.

[0139] f t = σ(W f h L (t) + U f h t-1 + b f ) (23)

[0140] (2) Input gate: Determines the amount of new information input at the current moment.

[0141] i t = σ(W i h L (t) + U i h t-1 + b i ) (24)

[0142] (3) Candidate memory cell: Generates new candidate memory values.

[0143] C′ t = tanh(W c h L (t) + Uc h t-1 +b c ) (25)

[0144] (4) Memory update: Update the memory cell at the current moment.

[0145] C t = f t ·C t-1 + i t ·C′ t (26)

[0146] (5) Output gate: Determine the output at the current moment.

[0147] o t = σ(W o h L (t)+ U o h t-1 + b o ) (27)

[0148] h t = o t ·tanh(C t ) (28)

[0149] Among them, h L (t) is the input data; h t-1 is the hidden state at the previous moment; c t is the cell state (memory cell) at the current moment; W (such as W f , W i , W c , W o ) are the weights of each gate, U (such as U f , U f , U c , U o ) are the weights of each state, b (such as b f , b i , b c , b o ) are the biases of each gate.

[0150] In multivariate prediction, LSTM not only processes multiple input variables but also needs to generate multiple output variables (i.e., multivariate regression). The output h t of the LSTM model can be passed to a fully connected (Dense) layer to generate the target prediction value.

[0151] Suppose q target variables are to be predicted. The last hidden state h T of the LSTM model will be passed to a fully connected layer to generate the predicted values of q target variables

[0152]

[0153] Among them, W out is the weight matrix of the output layer, with a size of q×h dim where h dim is the dimension of the LSTM hidden layer, and b out is the bias term of the output layer.

[0154] It should be noted that the LSTM model can be pre-trained. Exemplarily, the LSTM model can be trained by the backpropagation algorithm to minimize the loss between the predicted value and the true value. The training of the LSTM model can use the mean squared error (MSE) as the loss function:

[0155]

[0156] where is the predicted value of the LSTM model, and y t is the true value.

[0157] In the above embodiment, through the sliding window analysis of the vital sign data and combined with the LSTM algorithm, the abnormal changes of the vital signs can be identified in advance, especially the gradually abnormal trends, such as the continuous increase in heart rate, the continuous decrease in blood oxygen saturation, etc. For example, it is possible to predict whether states such as fatigue and abnormal breathing may occur within the next 5 minutes, and design a composite early warning model by combining the multi-dimensional prediction of vital sign data and environmental data (such as the changes in CO concentration and heart rate). Optionally, an early warning is issued only when multiple variables exceed the threshold to reduce false alarms.

[0158] It should be noted that in some embodiments, the effect of the intelligent prediction and early warning of the safety risks of underground personnel can be continuously monitored, and the DBN model and / or the vital sign change trend prediction model can be corrected according to the actual situation. Exemplarily, the effect of the intelligent prediction and early warning of the safety risks of underground personnel can be continuously monitored, and online learning and updating of the DBN model and / or the vital sign change trend prediction model can be performed according to the data feedback in actual use to adjust the sensitivity of the model. For example, the model parameters in the vital sign change trend prediction model can be adjusted, and / or the weight parameters in the comprehensive vital sign evaluation function can be adjusted, and / or the model parameters in the classification model can be adjusted, etc., so as to reduce false alarms and missed alarms and ensure the adaptability and accuracy of the model under different environmental conditions. Exemplarily, classification can be performed using the classification model, and the predicted vital sign change trend and environmental factor data are used as the comprehensive vital sign evaluation function Z nThe input is compared and classified with a set threshold for the output evaluation score (which is just a numerical value) to determine the warning level. Then, an ROC (Receiver Operating Characteristic curve) is plotted and the AUC (Area Under Curve, the area enclosed by the ROC curve and the coordinate axes) value is calculated to adjust the threshold to optimize the classification model.

[0159] Figure 4 This is a schematic flowchart of the intelligent prediction and early warning device for underground personnel safety risks provided by the embodiments of the present disclosure. As Figure 4 shown, the intelligent prediction and early warning device for underground personnel safety risks may include: an acquisition module 401, a prediction module 402, a determination module 403, a detection module 404, and an early warning module 405.

[0160] Among them, the acquisition module 401 is used to acquire the vital sign information and working environment information of underground workers within the first time period.

[0161] The prediction module 402 is used to predict the changing trend of the vital signs of the workers based on the vital sign information and the working environment information by using a pre-trained prediction model for the changing trend of vital signs, and obtain the changing trend of the vital signs of the workers. In some embodiments, the prediction module 402 is used to: based on the vital sign information and the working environment information, use the sliding window technique to construct a time series with a preset time length for the window, and the time series includes the vital sign information and the working environment information within multiple time windows; based on the pre-trained deep belief network DBN model, perform deep feature extraction on the time series to obtain a deep feature representation; the DBN model is composed of multiple restricted Boltzmann machines RBMs, the number of layer nodes of the multi-layer RBM structure in the DBN model is optimized based on the particle swarm optimization PSO algorithm, and each layer of the RBM in the DBN model is unsupervised trained based on the optimized number of nodes; according to the deep feature representation and the pre-trained prediction model for the changing trend of vital signs, predict the changing trend of the vital signs of the workers, and obtain the changing trend of the vital signs of the workers.

[0162] A determination module 403, configured to determine comprehensive vital sign assessment information of an operator according to the vital sign change trend and the operation environment information. In some embodiments, the determination module 403 is configured to: determine the maximum survival period of the human body corresponding to the vital sign change trend based on a preset quantitative relationship between the vital sign indicators and the maximum survival period of the human body; wherein, the quantitative relationship is established based on the multiple regression analysis method, the vital sign indicators are used as independent variables, and the maximum survival period of the human body is used as the dependent variable; perform weighted summation according to the operation environment information and its weight to obtain environmental risk score information; the weight of the operation environment information is determined based on the entropy weight method; and obtain the comprehensive vital sign assessment information of the operator by using a preset comprehensive vital sign assessment function according to the maximum survival period of the human body and the environmental risk score information.

[0163] A detection module 404, configured to detect an emergency according to the vital sign information and obtain a detection result for the emergency.

[0164] An early warning module 405, configured to perform hierarchical early warning according to the comprehensive vital sign assessment information of the operator, the vital sign change trend, and the detection result for the emergency, and obtain the safety risk early warning level of the operator.

[0165] In some embodiments, the early warning module 405 is configured to: classify the predicted vital sign states according to the vital sign change trend to obtain vital sign state categories; and perform hierarchical early warning based on the discrimination conditions associated with a set of multiple early warning levels, the comprehensive vital sign assessment information of the operator, the vital sign state categories, and the detection result for the emergency.

[0166] In some embodiments, the early warning module 405 is configured to: classify the predicted vital sign states based on a preset state assessment function and the vital sign change trend to obtain vital sign state categories; wherein, the state assessment function includes multiple state categories, at least one discrimination function corresponding to the multiple state categories, and at least one discrimination threshold, the at least one discrimination function is a function for modeling the vital sign change trend corresponding to each state category based on medical knowledge, and the discrimination threshold is used to determine the vital sign state category corresponding to the vital sign change trend.

[0167] In some embodiments, the early warning module 405 is configured to: classify the predicted vital sign states by using a classification algorithm based on the vital sign change trend to obtain vital sign state categories.

[0168] In some embodiments, the early warning module 405 is further configured to: trigger an early warning signal when the safety risk early warning level meets the early warning trigger mechanism; and / or provide intelligent decision support suggestions according to the safety risk early warning level.

[0169] Regarding the device in the above embodiments, the specific manners in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0170] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device and a readable storage medium.

[0171] As Figure 5 shown, it is a block diagram of an electronic device according to an embodiment of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as, a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, a personal digital processor, a cellular phone, a smart phone, a wearable device, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or claimed herein.

[0172] As Figure 5 shown, the electronic device includes: one or more processors 501, a memory 502, and an interface for connecting the components, including a high-speed interface and a low-speed interface. Each component is interconnected using different buses and can be installed on a common motherboard or in other ways as needed. The processor can process instructions executed within the electronic device, including instructions stored in the memory or on the memory to display graphical information of a GUI on an external input / output device (such as a display device coupled to the interface). In other embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple electronic devices can be connected, and each device provides some necessary operations (such as, as a server array, a set of blade servers, or a multi-processor system). Figure 5 In

[0173] The memory 502 is the non-transitory computer-readable storage medium provided by the present application. Among them, the memory stores instructions executable by at least one processor, so that the at least one processor executes the intelligent prediction and early warning method for underground personnel safety risks provided by the present application. The non-transitory computer-readable storage medium of the present application stores computer instructions, and the computer instructions are used to cause a computer to execute the intelligent prediction and early warning method for underground personnel safety risks provided by the present application.

[0174] The memory 502 serves as a non-transitory computer-readable storage medium and can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the intelligent prediction and early warning method for the safety risks of underground personnel in the embodiments of the present application (for example, the acquisition module 401, the determination module 402, the prediction module 403, the detection module 404, and the early warning module 405 shown in Figure 4 Figure). The processor 501 executes various functional applications and data processing of the server by running the non-transitory software programs, instructions, and modules stored in the memory 502, that is, implements the intelligent prediction and early warning method for the safety risks of underground personnel in the above method embodiments.

[0175] The memory 502 may include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the electronic device, etc. In addition, the memory 502 may include high-speed random access memory and may also include non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some embodiments, the memory 502 may optionally include a memory remotely provided relative to the processor 501, and these remote memories can be connected to the electronic device through a network. Examples of the above network include but are not limited to the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0176] The electronic device may further include: an input device 503 and an output device 504. The processor 501, the memory 502, the input device 503, and the output device 504 can be connected through a bus or other means, Figure 5 taking the connection through the bus as an example.

[0177] The input device 503 can receive input digital or character information and generate key signal inputs related to the user settings and function controls of the electronic device, such as input devices like a touch screen, a keypad, a mouse, a trackpad, a touchpad, a pointing stick, one or more mouse buttons, a trackball, a joystick, etc. The output device 504 may include a display device, an auxiliary lighting device (for example, an LED), and a tactile feedback device (for example, a vibration motor), etc. The display device may include but is not limited to a liquid crystal display (LCD), a light-emitting diode (LED) display, and a plasma display. In some embodiments, the display device may be a touch screen.

[0178] The various embodiments of the systems and techniques described herein can be implemented in digital electronic circuitry, integrated circuit systems, ASICs (application specific integrated circuits), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be a special-purpose or general-purpose programmable processor that receives data and instructions from, and transmits data and instructions to, a storage system, at least one input device, and at least one output device.

[0179] These computing programs (also referred to as programs, software, software applications, or code) include machine instructions for a programmable processor and can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and / or device (e.g., a magnetic disk, optical disk, memory, programmable logic device (PLD)) used to provide machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal used to provide machine instructions and / or data to a programmable processor.

[0180] For providing interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic, speech, or tactile input).

[0181] The systems and techniques described herein can be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected to each other by digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), the Internet, and blockchain networks.

[0182] A computer system can include a client and a server. The client and the server are generally far from each other and typically interact through a communication network. The client-server relationship is created by computer programs that run on the respective computers and have a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system, and solves the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services ("Virtual Private Server", or simply "VPS"). The server can also be a server of a distributed system, or a server combined with blockchain.

[0183] In the description of this specification, the description referring to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present disclosure. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0184] In the description of the present disclosure, the meaning of "at least one" is one or more, and the meaning of "a plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0185] Any process or method description represented in a flowchart or otherwise described herein can be understood to represent a module, segment, or portion of code including one or more executable instructions for implementing a customized logic function or process. The scope of the preferred embodiments of the present disclosure includes additional implementations in which functions may be executed not in the order shown or discussed, including in a substantially simultaneous manner according to the associated functions or in a reverse order, which should be understood by those skilled in the art to which the embodiments of the present disclosure pertain.

[0186] The logic and / or steps represented in a flowchart or otherwise described herein, for example, can be considered a sequenced list of executable instructions for implementing a logical function and can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection having one or more wires (electronic device), a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable medium on which the program can be printed, as the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpretation, or otherwise processing as appropriate, and then stored in a computer memory.

[0187] It should be understood that various parts of the present disclosure can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one of the following techniques known in the art or a combination thereof can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0188] Those of ordinary skill in the art can understand that all or part of the steps carried out in the method of the above embodiments can be completed by instructing relevant hardware through a program. The program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiments.

[0189] In addition, in each of the various embodiments of the present disclosure, each functional unit may be integrated in a processing module, or each unit may exist physically alone, or two or more units may be integrated in a module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0190] The above-mentioned storage medium may be a read-only memory, a magnetic disk, an optical disc, etc. Although the embodiments of the present disclosure have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present disclosure. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present disclosure.

Claims

1. An intelligent prediction and early warning method for underground personnel safety risks, characterized in that: include: Obtain the vital signs and working environment information of underground workers in the first place; According to the vital sign information and the working environment information, a pre-trained vital sign change trend prediction model is used to predict the vital sign change trend of the operator to obtain the vital sign change trend of the operator; Determining comprehensive assessment information of the vital signs of the operator according to the vital signs change trend and the working environment information; Detecting an emergency event according to the vital sign information to obtain a detection result for the emergency event; According to the comprehensive assessment information of the vital signs of the operator, the trend of the vital signs changes and the detection results of the emergency, a graded warning is performed to obtain the safety risk warning level of the operator.

2. The method according to claim 1, characterized in that The step of providing a graded warning based on the comprehensive evaluation information of the vital signs of the operator, the trend of the vital signs changes and the detection results of the emergency event includes: Classifying the predicted vital sign status according to the vital sign change trend to obtain a vital sign status category; Based on the set judgment conditions associated with multiple warning levels, the comprehensive assessment information of the vital signs of the operator, the vital signs status category and the detection results of the emergency, a graded warning is performed.

3. The method according to claim 2, characterized in that The predicted vital sign status is classified according to the vital sign change trend to obtain the vital sign status category, including: Based on a plurality of preset state evaluation functions and the trend of the vital signs change, the predicted vital signs state is classified to obtain the vital signs state category; the vital signs state category is used to determine whether the operator is in a dangerous state; Each of the state evaluation functions represents a mapping relationship, and the mapping relationship is a mapping relationship between a vital sign change trend and a vital sign state category.

4. The method according to claim 2, characterized in that The predicted vital sign status is classified according to the vital sign change trend to obtain the vital sign status category, including: Based on the changing trend of the vital signs, a classification algorithm is used to classify the predicted vital signs status to obtain the vital signs status category.

5. The method according to claim 1, characterized in that The method further comprises: When the security risk warning level meets the warning trigger mechanism, triggering a warning signal; and / or, Provide intelligent decision support suggestions based on the security risk warning level.

6. The method according to claim 1, characterized in that Determining the comprehensive assessment information of the vital signs of the operator according to the vital sign change trend and the working environment information includes: Based on the quantitative relationship between the preset vital sign indicators and the maximum survival period of the human body, determine the maximum survival period of the human body corresponding to the change trend of the vital signs; wherein the quantitative relationship is established based on a multiple regression analysis method, with the vital sign indicators as independent variables and the maximum survival period of the human body as the dependent variable; Performing weighted summation according to the operating environment information and its weight to obtain environmental risk score information; the weight of the operating environment information is determined based on an entropy weight method; According to the maximum survival period of the human body and the environmental risk score information, a preset vital sign comprehensive evaluation function is used to obtain the vital sign comprehensive evaluation information of the operator.

7. The method according to claim 1, characterized in that The method of predicting the vital sign change trend of the operator using a pre-trained vital sign change trend prediction model according to the vital sign information and the working environment information to obtain the vital sign change trend of the operator includes: Based on the vital sign information and the working environment information, a sliding window technology is used to construct a time series with a window of preset duration, wherein the time series includes the vital sign information and the working environment information within a plurality of time windows; Based on a pre-trained deep belief network DBN model, deep feature extraction of the time series data is performed to obtain a deep feature representation; the DBN model is composed of multiple restricted Boltzmann machines (RBMs), the number of layer nodes of the multi-layer RBM structure in the DBN model is optimized based on a particle swarm optimization (PSO) algorithm, and each layer of the RBM of the DBN model is unsupervisedly trained based on the optimized number of nodes; The vital signs change trend of the operator is predicted based on the deep feature representation and the pre-trained vital signs change trend prediction model to obtain the vital signs change trend of the operator.

8. An intelligent prediction and early warning device for underground personnel safety risks, characterized in that: include: An acquisition module is used to obtain the vital signs information and working environment information of underground workers in the first time; A prediction module, used to predict the vital sign change trend of the operator according to the vital sign information and the working environment information by using a pre-trained vital sign change trend prediction model to obtain the vital sign change trend of the operator; A determination module, used to determine the comprehensive assessment information of the vital signs of the operator according to the vital signs change trend and the working environment information; A detection module, used to detect an emergency event according to the vital sign information, and obtain a detection result for the emergency event; The early warning module is used to perform graded early warning according to the comprehensive evaluation information of the vital signs of the operator, the trend of the vital signs changes and the detection results of the emergency, so as to obtain the safety risk early warning level of the operator.

9. An electronic device, characterized in that: include: one or more processors; Wherein, the electronic device is used to execute the intelligent prediction and early warning method for underground personnel safety risks as described in any one of claims 1-7.

10. A storage medium storing instructions, characterized in that: When the instruction is executed on an electronic device, the electronic device executes the intelligent prediction and early warning method for underground personnel safety risks according to any one of claims 1 to 7.

Images