Cave microenvironment intelligent early warning system and method thereof

Abstract

The invention relates to a cave microenvironment intelligent early warning system and method. The cave microenvironment intelligent early warning system comprises an acquisition terminal, a monitoring terminal and an alarm terminal, the acquisition terminal sends the environmental parameter values acquired by the sensors of different types to the monitoring terminal; the monitoring terminal takes the environmental parameter values of the various types of sensors within the first preset duration inquired in a reverse order from the current acquisition moment as target environmental parameter values; according to the target environment parameter value of each type of sensor and a plurality of historical record environment parameter values pre-stored by the same type of sensor, predicting a predicted environment parameter value within a second preset time length in the future; taking the plurality of sensors with the predicted environment parameter values exceeding the corresponding preset parameter value threshold values as target sensors, and determining alarm type information according to the types of the target sensors; alarm state information is determined according to the predicted environment parameter values corresponding to the multiple target sensors, and an early warning instruction is generated and sent to an alarm terminal according to the alarm type information and the alarm state information; and the alarm terminal performs early warning according to the early warning instruction.

Description

Technical Field

[0001] The present disclosure relates to the technical field of environmental monitoring, and in particular to a cave microenvironment intelligent early warning system and method thereof. Background Art

[0002] The cave environment usually refers to the unique natural environment and ecosystem inside the cave, which is closed, complex and fragile. Changes in its microenvironmental parameters (such as temperature, humidity, gas concentration, air pressure, etc.) may have serious impacts on the cave ecosystem, archaeological relics or underground facilities. Traditional cave environment monitoring methods usually rely on single sensors and manual analysis, which have problems such as incomplete data, slow response speed and low warning accuracy. Therefore, there is an urgent need for an intelligent cave microenvironment early warning system that can monitor in real time, accurately analyze and warn. Summary of the invention

[0003] The purpose of the present invention is to provide a cave microenvironment intelligent early warning system and method thereof, aiming to improve the accuracy of cave environment detection and early warning.

[0004] In order to achieve the above-mentioned purpose, a first aspect of the embodiment of the present disclosure provides a cave microenvironment intelligent early warning system, the system comprising:

[0005] A collection terminal, a monitoring terminal connected to the collection terminal in communication, and an alarm terminal connected to the monitoring terminal in communication;

[0006] The acquisition terminal is configured with a plurality of sensors for collecting different types of environmental parameters. The acquisition terminal is used to be placed in the cave microenvironment, and to collect environmental parameter values ​​of corresponding environmental parameters in the cave microenvironment through a plurality of different types of sensors, and to send the environmental parameter values ​​collected by the different types of sensors to the monitoring terminal in chronological order;

[0007] The monitoring terminal is used to use the environmental parameter values ​​collected by the sensors of each type at each collection moment in a historical time period of a first preset time length in reverse order from the current collection moment as target environmental parameter values;

[0008] According to the target environmental parameter value corresponding to the sensor of each type at the current collection moment and a plurality of historical record environmental parameter values ​​pre-stored by the sensor of the same type, predict the predicted environmental parameter value of the environmental parameter corresponding to the sensor of the type within a future time period of a second preset time length starting from the current collection moment;

[0009] Taking multiple sensors whose predicted environmental parameter values ​​exceed the preset parameter value threshold of the corresponding type as target sensors, determining the alarm type information for the cave microenvironment according to the type corresponding to the target sensor; determining the alarm status information for the cave microenvironment according to the predicted environmental parameter values ​​corresponding to the multiple target sensors, and generating and sending an early warning instruction to the alarm terminal according to the alarm type information and the alarm status information;

[0010] The alarm terminal is used to receive the early warning instruction sent by the monitoring terminal, and issue an early warning according to the alarm type information and alarm status information carried in the early warning instruction.

[0011] In a possible implementation, the monitoring terminal is specifically used for:

[0012] Calculate the Pearson correlation coefficient between the target environmental parameter of each type and the multiple historical record environmental parameter values ​​pre-stored by the sensor of the same type according to the target environmental parameter value corresponding to the sensor of each type at the current acquisition moment and the multiple historical record environmental parameter values ​​pre-stored by the sensor of the same type;

[0013] Determine target historical environmental data from a plurality of historical record environmental parameter values ​​corresponding to the type according to the Pearson correlation coefficient corresponding to the target environmental parameter value of each type;

[0014] When the warning result information associated with the target historical environmental data corresponding to any type of sensor represents that an alarm event occurs within a future time period of the second preset time length in sequence, the predicted environmental parameter value of the environmental parameter corresponding to the sensor of this type in the future time period of the second preset time length starting from the current collection moment is predicted based on the target environmental parameter value collected by the sensor of this type at each collection moment in the historical time period of the first preset time length.

[0015] In a possible implementation, the monitoring terminal is specifically used for:

[0016] According to the target environmental parameter values ​​collected by the sensor of this type at each collection moment in the historical time period of the first preset time period, the target environmental parameter values ​​in the historical time period of the first preset time period are divided into multiple sets corresponding to the first random number through multiple random time periods of a first random number, wherein each set obtained by the first random number includes a second random number of target environmental parameter values ​​at continuous collection times of any of the random time periods;

[0017] According to the order in which the sets appear in the historical time period of the first preset time length, the target environment parameter values ​​included in the previous set are traversed, and the predicted parameter values ​​of the subsequent set are predicted by the deep learning model, and the model parameters of the deep learning model are adjusted according to the error between the target environment parameter value of the subsequent set and the predicted parameter value predicted by the previous set, wherein the target environment parameter value corresponding to each previous set is input into the deep learning model after the model parameter of the previous set is adjusted as the subsequent set;

[0018] After the set that is last in the historical time period of the first preset duration is used as the subsequent set for model parameter adjustment, the target environmental parameter values ​​collected by the sensor of this type at each collection moment in the historical time period of the first preset duration are input into the deep learning model after the final model parameter adjustment to obtain the predicted parameter values ​​output by the deep learning model, and to obtain the predicted environmental parameter values ​​of the environmental parameters corresponding to the sensor of this type in the future time period of the second preset duration starting from the current collection moment.

[0019] In a possible implementation, the monitoring terminal is specifically used for:

[0020] For any of the types of the sensors, randomly determine a first random number of the sets and randomly determine a second random number of the target environmental parameter values ​​included in each of the sets;

[0021] Constructing a plurality of the sets according to the first random number, and adding an index to each of the sets in sequence according to time;

[0022] According to the index of each of the sets and the second random number of the target environmental parameter values ​​included in each of the sets, the target environmental parameter values ​​within the historical time period of the first preset time length are divided into each of the sets, thereby completing the division of the target environmental parameter values ​​within the historical time period of the first preset time length into multiple sets corresponding to the first random number.

[0023] In a possible implementation, the monitoring terminal is specifically used for:

[0024] Calculate the first parameter value mean of the target environmental parameter value of each type of sensor according to the parameter value corresponding to each collection moment in the target environmental parameter value corresponding to each type of sensor, and calculate the second parameter value mean corresponding to each of the historical record environmental parameter values ​​pre-stored by the same type of sensor according to the parameter value corresponding to each collection moment in a plurality of historical record environmental parameter values ​​pre-stored by the same type of sensor;

[0025] Calculate, according to the mean value of the first parameter values ​​corresponding to each type of sensor and the parameter values ​​corresponding to each collection moment in the target environment parameter values ​​corresponding to the corresponding type of sensor, a first deviation between each parameter value in the target environment parameter values ​​of each type of sensor and the mean value of the first parameter values;

[0026] Calculate, according to the average of the second parameter values ​​corresponding to each of the historical record environmental parameter values ​​corresponding to each of the types of sensors and the parameter values ​​corresponding to each of the collection moments in the historical record environmental parameter values, a second deviation between each of the parameter values ​​in the historical record environmental parameter values ​​and the average of the second parameter values;

[0027] According to the first deviation and the second deviation corresponding to each type of sensor, after calculating the product of the first deviation and the second deviation corresponding to the type of sensor, summing the products to obtain a deviation product sum corresponding to the type of sensor;

[0028] Calculating the sum of squares of the first deviations corresponding to each type of sensor to obtain a first sum of squares of the deviations, and calculating the sum of squares of the second deviations corresponding to each type of sensor to obtain a second sum of squares of the deviations;

[0029] The Pearson correlation coefficient between the target environmental parameter value of each type of sensor and each of the historical record environmental parameter values ​​pre-stored by the same type of sensor is calculated based on the deviation product sum, the first deviation square sum and the second deviation square sum corresponding to each type of sensor.

[0030] In a possible implementation, the monitoring terminal is specifically used for:

[0031] The historical record environment parameter value whose Pearson correlation coefficient of each type of sensor is greater than or equal to the preset coefficient threshold corresponding to the corresponding type of sensor is used as a candidate historical record environment parameter value;

[0032] Calculate the mean value of the Pearson correlation coefficient corresponding to the candidate historical record environmental parameter value corresponding to each type of sensor;

[0033] Counting the number of candidate historical record environmental parameter values ​​corresponding to each type of sensor;

[0034] When the number of candidate historical record environment parameter values ​​corresponding to any type of sensor is greater than or equal to a preset number threshold of the corresponding type of sensor, the candidate historical record environment parameter value whose Pearson correlation coefficient in the type of sensor is greater than or equal to the corresponding mean is determined as the target historical environment data of the type of sensor.

[0035] In a possible implementation, the monitoring terminal is further used for:

[0036] Constructing a collection interval duration cost matrix based on energy consumption cost and accuracy cost according to a preset collection period set corresponding to the target sensor, an environmental parameter value of the target sensor at the current collection time, and the predicted environmental parameter value;

[0037] According to the adjustment cost matrix, multiple collection interval durations of the target sensor in the future time period starting from the current collection moment are determined, and according to the multiple collection interval durations corresponding to the multiple target sensors, the target collection interval durations of the multiple target sensors within the second preset duration are determined.

[0038] In a possible implementation, the sensors for collecting different types of environmental parameters include multiple ones of the following: a temperature sensor, a relative humidity sensor, a gas sensor, an air pressure sensor, a positive and negative ion sensor, and a light sensor.

[0039] A second aspect of the embodiments of the present disclosure provides a cave microenvironment intelligent early warning method, which is applied to a monitoring terminal in a cave microenvironment intelligent early warning system as described in any one of the first aspects, wherein the cave microenvironment intelligent early warning system includes a collection terminal and an alarm terminal that are communicatively connected to the monitoring terminal, and the method includes:

[0040] Receiving the environmental parameter values ​​of the corresponding environmental parameters in the cave microenvironment collected by the multiple different types of sensors sent by the collection terminal in time sequence, wherein the collection terminal is placed in the cave microenvironment, and collects the environmental parameter values ​​of the corresponding environmental parameters in the cave microenvironment through the multiple different types of sensors;

[0041] The environmental parameter values ​​collected by the sensors of each type at each collection moment in a historical time period of a first preset length of time in reverse order from the current collection moment are used as target environmental parameter values;

[0042] According to the target environmental parameter value corresponding to the sensor of each type at the current collection moment and a plurality of historical record environmental parameter values ​​pre-stored by the sensor of the same type, predict the predicted environmental parameter value of the environmental parameter corresponding to the sensor of the type within a future time period of a second preset time length starting from the current collection moment;

[0043] Taking multiple sensors whose predicted environmental parameter values ​​exceed the preset parameter value thresholds of corresponding types as target sensors, and determining the alarm type information for the cave microenvironment according to the types corresponding to the target sensors;

[0044] According to the predicted environmental parameter values ​​corresponding to the multiple target sensors, the alarm status information for the cave microenvironment is determined, and according to the alarm type information and the alarm status information, an early warning instruction is generated and sent to the alarm terminal, and the early warning instruction is used for the alarm terminal to issue an early warning according to the alarm type information and the alarm status information carried in the early warning instruction.

[0045] In a possible implementation, predicting the predicted environmental parameter value of the environmental parameter corresponding to the sensor of each type at the current collection moment and a plurality of historical record environmental parameter values ​​pre-stored by the sensor of the same type within a future time period of a second preset time length from the current collection moment includes:

[0046] Calculate the Pearson correlation coefficient between the target environmental parameter of each type and the multiple historical record environmental parameter values ​​pre-stored by the sensor of the same type according to the target environmental parameter value corresponding to the sensor of each type at the current acquisition moment and the multiple historical record environmental parameter values ​​pre-stored by the sensor of the same type;

[0047] Determine target historical environmental data from a plurality of historical record environmental parameter values ​​corresponding to the type according to the Pearson correlation coefficient corresponding to the target environmental parameter value of each type;

[0048] When the warning result information associated with the target historical environmental data corresponding to any type of sensor represents that an alarm event occurs within a future time period of the second preset time length in sequence, the predicted environmental parameter value of the environmental parameter corresponding to the sensor of this type in the future time period of the second preset time length starting from the current collection moment is predicted based on the target environmental parameter value collected by the sensor of this type at each collection moment in the historical time period of the first preset time length.

[0049] The present invention provides a cave microenvironment intelligent early warning system and method thereof. Compared with the prior art, it has the following beneficial effects:

[0050] By configuring multiple sensors of different types on the acquisition terminal, the key parameters in the cave microenvironment can be fully covered to ensure the comprehensiveness and accuracy of data collection. The high precision and diversity of the sensors help capture subtle changes in the cave environment. The monitoring terminal reversely queries the environmental parameter values ​​at each acquisition moment in the historical time period of the first preset time, which helps to identify the changing trend and periodicity of the environmental parameters and improve the accuracy of the prediction. Combining the current moment with historical data and using advanced data analysis algorithms for prediction, it is possible to gain insight into potential environmental risks in advance and provide a time window for early warning. According to the comparison between the predicted environmental parameter values ​​and the preset parameter value thresholds, the target sensors are automatically identified and screened, thereby improving the accuracy of determining the alarm type information and alarm status information, accurately reflecting the specific problems and severity of the cave environment, and the generation and sending of early warning instructions realize the instant transmission of information, ensuring that relevant personnel can quickly take countermeasures to reduce the possibility of disasters or mitigate their impact. The alarm terminal performs corresponding early warning operations according to the early warning instructions, which enhances the practicality and reliability of the system. In summary, by integrating high-precision sensors, intelligent data analysis algorithms and efficient information flow mechanisms, the accuracy, timeliness and practicality of cave environment detection and early warning are significantly improved.

[0051] Other features and advantages of the present disclosure will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The accompanying drawings are used to provide a further understanding of the present disclosure and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the present disclosure but do not constitute a limitation of the present disclosure. In the accompanying drawings:

[0053] Figure 1 It is a block diagram of a cave microenvironment intelligent early warning system according to an embodiment of the specification.

[0054] Figure 2 It is a flow chart of a cave microenvironment intelligent early warning method according to an embodiment of the specification.

[0055] Figure 3 It is a block diagram of a monitoring terminal according to an embodiment of the specification. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0057] The specific implementation of the present disclosure is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described herein is only used to illustrate and explain the present disclosure, and is not used to limit the present disclosure.

[0058] The present invention provides a cave microenvironment intelligent early warning system. Figure 1 As shown, the system comprises:

[0059] A collection terminal, a monitoring terminal connected to the collection terminal in communication, and an alarm terminal connected to the monitoring terminal in communication;

[0060] The acquisition terminal is deployed at the monitoring site and is responsible for real-time acquisition of various environmental parameters in the cave microenvironment. It integrates a variety of sensors, can automatically and continuously collect environmental data, and transmit these data to the monitoring center or other related equipment through communication.

[0061] The monitoring terminal is a central device or software system that receives, processes and displays environmental parameter data from the acquisition terminal. It is responsible for data storage, analysis, prediction, and the generation and sending of early warning instructions.

[0062] The alarm terminal is a device or system that receives the warning instructions sent by the monitoring terminal and performs warning operations according to the alarm type information and alarm status information in the instructions. It can be an audible and visual alarm, a text message notification system, an automatically activated emergency device, etc.

[0063] The acquisition terminal is configured with a plurality of sensors for collecting different types of environmental parameters. The acquisition terminal is used to be placed in the cave microenvironment, and to collect environmental parameter values ​​of corresponding environmental parameters in the cave microenvironment through a plurality of different types of sensors, and to send the environmental parameter values ​​collected by the different types of sensors to the monitoring terminal in chronological order;

[0064] In the disclosed embodiment, a plurality of high-precision sensors are configured inside the acquisition terminal, which can sense different physical and chemical parameters (such as temperature, humidity, gas concentration, etc.) in the cave microenvironment. When the acquisition terminal is placed in the cave, these sensors start working and collect environmental parameter values ​​in real time. The collected data will be packaged in chronological order and sent to the monitoring terminal via wired or wireless communication technology (such as Wi-Fi, Bluetooth, LoRa, etc.).

[0065] The monitoring terminal is used to use the environmental parameter values ​​collected by the sensors of each type at each collection moment in a historical time period of a first preset time length in reverse order from the current collection moment as target environmental parameter values;

[0066] According to the target environmental parameter value corresponding to the sensor of each type at the current collection moment and a plurality of historical record environmental parameter values ​​pre-stored by the sensor of the same type, predict the predicted environmental parameter value of the environmental parameter corresponding to the sensor of the type within a future time period of a second preset time length starting from the current collection moment;

[0067] Taking multiple sensors whose predicted environmental parameter values ​​exceed the preset parameter value threshold of the corresponding type as target sensors, determining the alarm type information for the cave microenvironment according to the type corresponding to the target sensor; determining the alarm status information for the cave microenvironment according to the predicted environmental parameter values ​​corresponding to the multiple target sensors, and generating and sending an early warning instruction to the alarm terminal according to the alarm type information and the alarm status information;

[0068] In the disclosed embodiment, after the cave microenvironment parameters are predicted, a series of predicted environmental parameter values ​​of sensors in the future time period are obtained. In order to determine which sensors may trigger an alarm, it is necessary to compare these predicted values ​​with preset parameter value thresholds. These thresholds are pre-set according to the characteristics and safety requirements of the cave microenvironment and are used to determine whether the environmental parameters are within a safe range.

[0069] Specifically, for each type of sensor, there is a corresponding preset parameter value threshold. When the predicted environmental parameter value of a sensor exceeds this threshold, the environmental parameter value corresponding to this sensor in the cave microenvironment may be abnormal and need to be paid attention to. Therefore, these sensors whose predicted values ​​exceed the threshold are used as target sensors, which may be the key factors leading to cave microenvironment alarms.

[0070] In the disclosed embodiment, after the target sensors are determined, the alarm type information is determined according to the types of these sensors. Different sensor types correspond to different environmental parameters, such as temperature, humidity, oxygen concentration, etc. Therefore, when a certain type of sensor is determined as a target sensor, the type of alarm that may occur can be determined based on this type.

[0071] For example, if the target sensor is a temperature sensor and its predicted value exceeds a preset high temperature threshold, then the alarm type can be determined as "high temperature alarm". Similarly, if the target sensor is an oxygen concentration sensor and its predicted value is lower than a preset low oxygen threshold, then the alarm type can be determined as "low oxygen alarm". In this way, the alarm type information can be quickly determined according to the type of the target sensor, providing strong support for subsequent alarm processing and emergency response.

[0072] In the disclosed embodiment, the alarm status information is determined according to the predicted environmental parameter values ​​corresponding to the multiple target sensors. The alarm status information is used to describe the overall safety status of the cave microenvironment and the degree of possible danger.

[0073] Specifically, the overall safety status of the cave microenvironment can be evaluated based on the number of target sensors and their corresponding predicted environmental parameter values. If the number of target sensors is large and their predicted values ​​seriously exceed the preset threshold, then the multi-sensor data can be fused through, for example, Kalman filtering or particle filtering, for mutual verification and rejection, to determine whether the cave microenvironment is in a serious dangerous state. On the contrary, if the number of target sensors is small and their predicted values ​​only slightly exceed the preset threshold, then it can be considered that the cave microenvironment is in a relatively safe state, but it still needs to be closely monitored.

[0074] In addition, the danger level of the cave microenvironment can be assessed based on the change trend of the predicted environmental parameter values ​​corresponding to the target sensor. If the predicted value shows a continuous upward or downward trend and exceeds the preset threshold range, it can be considered that the danger level is high and intervention measures need to be taken as soon as possible.

[0075] In summary, by determining the target sensor, alarm type information and alarm status information, real-time monitoring and early warning of the cave microenvironment can be achieved, thereby providing a strong safety guarantee for activities such as cave exploration and scientific research, and helping to reduce potential environmental risks.

[0076] In the disclosed embodiment, after the monitoring terminal receives the environmental parameter data sent by the collection terminal, it will first store the data in chronological order to form a historical database. Then, it will reversely query the historical data within the first preset time period from the current collection moment as the target environmental parameter value. Next, the monitoring terminal will use advanced data analysis algorithms (such as time series analysis, machine learning models, etc.) and pre-stored multiple historical record environmental parameter values ​​to predict the environmental parameters at the current collection moment, and obtain the predicted environmental parameter values ​​within the second preset time period in the future. These predicted values ​​will be compared with the preset parameter value threshold to determine whether there is a potential environmental risk.

[0077] The alarm terminal is used to receive the early warning instruction sent by the monitoring terminal, and issue an early warning according to the alarm type information and alarm status information carried in the early warning instruction.

[0078] In the disclosed embodiment, when the monitoring terminal detects that the predicted environmental parameter value exceeds the preset threshold, an early warning instruction is generated and sent to the alarm terminal via communication. After receiving the instruction, the alarm terminal will perform the corresponding early warning operation according to the alarm type information and alarm status information in the instruction. For example, if it refers to an alarm of too low oxygen concentration, the alarm terminal may start the oxygen supplement equipment; if it refers to an alarm of excessive harmful gas, it may trigger an audible and visual alarm and send a text message to notify relevant personnel. These early warning operations are intended to promptly remind relevant personnel to pay attention and take necessary emergency measures to prevent or mitigate potential environmental risks.

[0079] The above technical solution can fully cover the key parameters in the cave microenvironment by configuring multiple different types of sensors on the acquisition terminal, ensuring the comprehensiveness and accuracy of data acquisition. The high precision and diversity of the sensors help capture subtle changes in the cave environment. The monitoring terminal reversely queries the environmental parameter values ​​at each acquisition moment in the historical time period of the first preset time, which helps to identify the changing trend and periodic law of the environmental parameters and improve the accuracy of the prediction. Combining the current moment with historical data and using advanced data analysis algorithms for prediction, it is possible to gain insight into potential environmental risks in advance and provide a time window for early warning. According to the comparison between the predicted environmental parameter values ​​and the preset parameter value thresholds, the target sensors are automatically identified and screened, thereby improving the accuracy of determining the alarm type information and alarm status information, accurately reflecting the specific problems and severity of the cave environment, and the generation and sending of early warning instructions realize the instant transmission of information, ensuring that relevant personnel can quickly take countermeasures to reduce the possibility of disasters or mitigate their impact. The alarm terminal performs corresponding early warning operations according to the early warning instructions, enhancing the practicality and reliability of the system. In summary, by integrating high-precision sensors, intelligent data analysis algorithms and efficient information flow mechanisms, the accuracy, timeliness and practicality of cave environment detection and early warning are significantly improved.

[0080] In a possible implementation, the monitoring terminal is specifically used for:

[0081] Calculate the Pearson correlation coefficient between the target environmental parameter of each type and the multiple historical record environmental parameter values ​​pre-stored by the sensor of the same type according to the target environmental parameter value corresponding to the sensor of each type at the current acquisition moment and the multiple historical record environmental parameter values ​​pre-stored by the sensor of the same type;

[0082] Among them, the Pearson Correlation Coefficient is a statistical indicator used to measure the degree of linear correlation between two variables. Its value is between -1 and 1, 1 represents a perfect positive correlation, -1 represents a perfect negative correlation, and 0 represents no correlation.

[0083] In the disclosed embodiment, the Pearson correlation coefficient is used to calculate the linear correlation between the target environmental parameter value at the current acquisition time and multiple historical environmental parameter values ​​pre-stored by the same type of sensor. Specifically, for each type of sensor, the monitoring terminal calculates the Pearson correlation coefficient between the target environmental parameter value at the current acquisition time and each historical environmental parameter value. These correlation coefficients reflect the similarity between the current environmental parameters and the historical environmental parameters and the consistency of the change trend.

[0084] Determine target historical environmental data from a plurality of historical record environmental parameter values ​​corresponding to the type according to the Pearson correlation coefficient corresponding to the target environmental parameter value of each type;

[0085] The target historical environmental data refers to historical data that is screened from multiple historical environmental parameter values ​​according to the Pearson correlation coefficient and has a high linear correlation with the current environmental parameters.

[0086] In the disclosed embodiment, when determining the target historical environmental data, the monitoring terminal will select several historical records with the highest correlation with the current environmental parameter value as the target historical environmental data according to the calculated Pearson correlation coefficient. These target historical environmental data will be used for subsequent prediction analysis because they have similar change trends and characteristics to the current environmental parameters and can provide a more accurate prediction basis.

[0087] When the warning result information associated with the target historical environmental data corresponding to any type of sensor represents that an alarm event occurs within a future time period of the second preset time length in sequence, the predicted environmental parameter value of the environmental parameter corresponding to the sensor of this type in the future time period of the second preset time length starting from the current collection moment is predicted based on the target environmental parameter value collected by the sensor of this type at each collection moment in the historical time period of the first preset time length.

[0088] Among them, the warning result information refers to the record information stored in association with the target historical environmental data, which indicates whether an alarm event will occur in the future time period. The prediction adjustment based on the warning result information refers to adjusting the prediction model or prediction process according to the warning result information to improve the accuracy of the prediction.

[0089] In the disclosed embodiment, in the monitoring terminal, historical warning result information is stored for each type of sensor. When predicting future environmental parameters, the monitoring terminal will first check the warning result information associated with the target historical environmental data. If this information indicates that an alarm event has occurred in a future time period of a second preset time period in sequence, the monitoring terminal will adjust the prediction model based on this information. Specifically, it will pay more attention to the changing trends and characteristics of environmental parameters that have caused alarm events in the historical time period, thereby giving higher weights when predicting future environmental parameters. This prediction adjustment method based on warning result information can make the prediction results closer to the actual situation and improve the accuracy and reliability of the warning.

[0090] In summary, by calculating the Pearson correlation coefficient, determining the target historical environmental data, and making prediction adjustments based on the early warning result information, this technology can further improve the prediction accuracy of cave microenvironment parameters, thereby providing more effective early warning and emergency response.

[0091] In a possible implementation, the monitoring terminal is specifically used for:

[0092] According to the target environmental parameter values ​​collected by the sensor of this type at each collection moment in the historical time period of the first preset time period, the target environmental parameter values ​​in the historical time period of the first preset time period are divided into multiple sets corresponding to the first random number through multiple random time periods of a first random number, wherein each set obtained by the first random number includes a second random number of target environmental parameter values ​​at continuous collection times of any of the random time periods;

[0093] Among them, random time division is to divide the historical time period of the first preset time period into multiple sub-time periods according to the randomly selected time period, and the data in each sub-time period constitutes a set. This division method helps to increase the diversity of data and the generalization ability of the model.

[0094] In the disclosed embodiment, the monitoring terminal divides the target environmental parameter values ​​within the historical time period of the first preset time period into multiple sets according to the first random number (i.e., the number of randomly selected sub-time periods) and the random duration of each sub-time period. Each set contains the target environmental parameter values ​​of the second random number of continuous collection times of any random duration. This division method ensures that the data in each set is both representative and random, which is helpful for the subsequent training and optimization of deep learning models.

[0095] According to the order in which the sets appear in the historical time period of the first preset time length, the target environment parameter values ​​included in the previous set are traversed, and the predicted parameter values ​​of the subsequent set are predicted by the deep learning model, and the model parameters of the deep learning model are adjusted according to the error between the target environment parameter value of the subsequent set and the predicted parameter value predicted by the previous set, wherein the target environment parameter value corresponding to each previous set is input into the deep learning model after the model parameter of the previous set is adjusted as the subsequent set;

[0096] In the disclosed embodiment, the monitoring terminal will traverse these sets in the order in which the sets appear in the historical time period of the first preset time. For each previous set, the monitoring terminal will input the target environmental parameter value therein into the deep learning model to predict the predicted parameter value of the subsequent set. Then, according to the error between the target environmental parameter value and the predicted parameter value of the subsequent set, the model parameters of the deep learning model are adjusted by the back propagation algorithm. This process will be iterated until all sets have undergone a model parameter adjustment as the subsequent set.

[0097] The target environment parameter values ​​corresponding to each previous set are input into the deep learning model again after the model parameters of the previous set are adjusted as the subsequent set. This iterative adjustment method helps the model gradually learn the inherent laws and characteristics of the data and improve the accuracy of the prediction.

[0098] After the set that is last in the historical time period of the first preset duration is used as the subsequent set for model parameter adjustment, the target environmental parameter values ​​collected by the sensor of this type at each collection moment in the historical time period of the first preset duration are input into the deep learning model after the final model parameter adjustment to obtain the predicted parameter values ​​output by the deep learning model, and to obtain the predicted environmental parameter values ​​of the environmental parameters corresponding to the sensor of this type in the future time period of the second preset duration starting from the current collection moment.

[0099] In the disclosed embodiment, after the model parameter adjustment of all sets is completed, the monitoring terminal will input the target environmental parameter values ​​collected at each collection moment in the historical time period of the first preset time into the deep learning model after the model parameter adjustment. Based on these input data, the model will output the predicted values ​​of the environmental parameters corresponding to this type of sensor in the future time period (i.e., the future time period of the second preset time period). These predicted values ​​will serve as the basis for early warning and emergency response, helping relevant personnel to take timely measures to deal with potential environmental risks.

[0100] In summary, through technical means such as random time division and set construction, deep learning model prediction and parameter adjustment, and final prediction and output, it is possible to accurately predict and timely warn of cave microenvironment parameters, providing more reliable safety guarantees for cave exploration, scientific research and other activities.

[0101] In a possible implementation, the monitoring terminal is specifically used for:

[0102] For any of the types of the sensors, randomly determine a first random number of the sets and randomly determine a second random number of the target environmental parameter values ​​included in each of the sets;

[0103] The first random number refers to the number of randomly selected sets, which is used to divide the target environmental parameter values ​​in the historical time period into multiple sets. The second random number refers to the number of target environmental parameter values ​​randomly included in each set.

[0104] In the disclosed embodiment, for any type of sensor, the monitoring terminal randomly determines the first random number of sets (i.e., the number of sets) and the second random number of target environmental parameter values ​​included in each set (i.e., the number of data in each set). The introduction of such randomness helps to increase the diversity of data and the generalization ability of the model, and prevents the model from overfitting.

[0105] Constructing a plurality of the sets according to the first random number, and sequentially adding an index to each of the sets according to time;

[0106] The set construction refers to the process of dividing the target environmental parameter values ​​in the historical time period into multiple sets according to the determined first random number and second random number. The index addition refers to assigning a unique identifier (i.e., index) to each set so that these sets can be traversed in order later.

[0107] In the disclosed embodiment, the monitoring terminal constructs a corresponding number of empty sets according to the determined first random number. Then, the target environmental parameter values ​​within the historical time period are divided into these sets in chronological order. During the division process, the monitoring terminal determines the amount of data to be included in each set according to the second random number. At the same time, a unique index value is assigned to each set so that these sets can be traversed and accessed in order later.

[0108] According to the index of each of the sets and the second random number of the target environmental parameter values ​​included in each of the sets, the target environmental parameter values ​​within the historical time period of the first preset time length are divided into each of the sets, thereby completing the division of the target environmental parameter values ​​within the historical time period of the first preset time length into multiple sets corresponding to the first random number.

[0109] In the disclosed embodiment, after the construction of the set and the addition of the index are completed, the monitoring terminal will divide the target environment parameter values ​​within the historical time period of the first preset time into each set according to the index of each set and the second random number of the target environment parameter values ​​included in each set. Specifically, the monitoring terminal will traverse the target environment parameter values ​​within the historical time period in chronological order, and fill the data into the corresponding set in sequence according to the index and capacity (i.e., the second random number) of each set. If a set has been filled with the second random number of data, continue to fill the next set until all target environment parameter values ​​are divided into corresponding sets.

[0110] Through the above steps, the monitoring terminal successfully divides the target environmental parameter values ​​within the historical time period of the first preset time into multiple sets corresponding to the first random number, and assigns a unique index value to each set. These sets will serve as the basic data for subsequent deep learning model training and prediction.

[0111] In summary, the random division and set construction of target environmental parameter values ​​in the historical time period are realized through technical means such as random number determination, set construction and index addition, data division and set filling.

[0112] In a possible implementation, the monitoring terminal is specifically used for:

[0113] Calculate the first parameter value mean of the target environmental parameter value of each type of sensor according to the parameter value corresponding to each collection moment in the target environmental parameter value corresponding to each type of sensor, and calculate the second parameter value mean corresponding to each of the historical record environmental parameter values ​​pre-stored by the same type of sensor according to the parameter value corresponding to each collection moment in a plurality of historical record environmental parameter values ​​pre-stored by the same type of sensor;

[0114] The first parameter value mean refers to the average value of the target environmental parameter values ​​corresponding to all acquisition times of a certain type of sensor in the current monitoring cycle. The second parameter value mean refers to the average value of the parameter values ​​corresponding to all acquisition times of each historical record environmental parameter value (i.e., data in a certain monitoring cycle in the past) in the historical records of the same type of sensor.

[0115] In the disclosed embodiment, for each type of sensor, the target environmental parameter values ​​at all acquisition moments in the current monitoring cycle are first extracted, and the arithmetic mean of these parameter values ​​is calculated to obtain the first parameter value mean. Then, for each historical record environmental parameter value of the same type of sensor, the arithmetic mean of the parameter values ​​at all acquisition moments in these records is also calculated to obtain the second parameter value mean. This step is intended to obtain the average performance of the sensor in different past monitoring cycles for comparison with the current monitoring cycle.

[0116] Calculate, according to the mean value of the first parameter values ​​corresponding to each type of sensor and the parameter values ​​corresponding to each collection moment in the target environment parameter values ​​corresponding to the corresponding type of sensor, a first deviation between each parameter value in the target environment parameter values ​​of each type of sensor and the mean value of the first parameter values;

[0117] Among them, the first deviation refers to the difference between the parameter value at each acquisition moment and the mean value of the first parameter value in the target environmental parameter value of a certain type of sensor. In the embodiment of the present disclosure, for each type of sensor, the difference between the parameter value at each acquisition moment in its target environmental parameter value and the mean value of the first parameter value is calculated to obtain the first deviation. This step is intended to measure the degree of deviation between the parameter value at each acquisition moment and the mean value in the current monitoring period.

[0118] Calculate, according to the average of the second parameter values ​​corresponding to each of the historical record environmental parameter values ​​corresponding to each of the types of sensors and the parameter values ​​corresponding to each of the collection moments in the historical record environmental parameter values, a second deviation between each of the parameter values ​​in the historical record environmental parameter values ​​and the average of the second parameter values;

[0119] The second deviation refers to the difference between the parameter value at each acquisition moment and the mean value of the second parameter value in a certain historical record of environmental parameter values ​​of the same type of sensor.

[0120] Similarly, for each historical environmental parameter value of the same type of sensor, the difference between the parameter value at each acquisition moment and the mean value of the second parameter value is calculated to obtain the second deviation. This step is intended to measure the degree of deviation between the parameter value at each acquisition moment and the mean value in the past monitoring period.

[0121] According to the first deviation and the second deviation corresponding to each type of sensor, after calculating the product of the first deviation and the second deviation corresponding to the type of sensor, summing the products to obtain a deviation product sum corresponding to the type of sensor;

[0122] The deviation product sum refers to the sum of the products of the first deviation of a certain type of sensor and the second deviation of the corresponding historical record.

[0123] In the embodiment of the present disclosure, for each type of sensor, its first deviation is multiplied by the second deviation of the corresponding historical record, and these products are summed to obtain the deviation product sum. This step is intended to measure the strength of the correlation between the current monitoring period and the historical record, and to reflect whether the change trends between the two are consistent in the form of a product.

[0124] Calculating the sum of squares of the first deviations corresponding to each type of sensor to obtain a first sum of squares of the deviations, and calculating the sum of squares of the second deviations corresponding to each type of sensor to obtain a second sum of squares of the deviations;

[0125] The sum of squared deviations refers to the sum of squares of the first deviation or second deviation of a certain type of sensor.

[0126] In the disclosed embodiment, at the same time, the sum of the squares of the first deviation and the sum of the squares of the second deviation of each type of sensor are calculated to obtain the first deviation square sum and the second deviation square sum, respectively. These values ​​are used to subsequently calculate the Pearson correlation coefficient to quantify the degree of linear correlation between the current monitoring period and the historical records.

[0127] The Pearson correlation coefficient between the target environmental parameter value of each type of sensor and each of the historical record environmental parameter values ​​pre-stored by the same type of sensor is calculated based on the deviation product sum, the first deviation square sum and the second deviation square sum corresponding to each type of sensor.

[0128] In the disclosed embodiment, based on the deviation product sum, the first deviation square sum and the second deviation square sum of each type of sensor, the Pearson correlation coefficient calculation formula (i.e., the correlation coefficient is equal to the deviation product sum divided by the product of the square root of the square sum of the deviations of the two variables) is used to calculate the Pearson correlation coefficient between the target environmental parameter value of the current monitoring period and the historical record environmental parameter value. This step is intended to quantify the degree of linear correlation between the two, so as to evaluate whether the data of the current monitoring period has a similar change trend and correlation with the historical record.

[0129] In summary, by calculating the mean, deviation, product sum of deviations, sum of squared deviations, and Pearson correlation coefficient of parameter values, this technology can realize in-depth analysis of changes in cave microenvironment parameters and provide a scientific basis for subsequent monitoring, early warning, and emergency response.

[0130] In a possible implementation, the monitoring terminal is specifically used for:

[0131] The historical record environment parameter value whose Pearson correlation coefficient of each type of sensor is greater than or equal to the preset coefficient threshold corresponding to the corresponding type of sensor is used as a candidate historical record environment parameter value;

[0132] In the disclosed embodiment, for each type of sensor, all historical records of environmental parameter values ​​are traversed, and historical records with a Pearson correlation coefficient greater than or equal to a preset coefficient threshold are screened out as candidate historical records of environmental parameter values. This step is intended to retain those historical records that have a high linear correlation with the current monitoring period data for subsequent analysis.

[0133] Calculate the mean value of the Pearson correlation coefficient corresponding to the candidate historical record environmental parameter value corresponding to each type of sensor;

[0134] The mean Pearson correlation coefficient refers to the arithmetic mean of the Pearson correlation coefficients corresponding to all candidate historical record environmental parameter values ​​of a certain type of sensor.

[0135] In the embodiment of the present disclosure, for each type of sensor, the arithmetic mean of the Pearson correlation coefficients corresponding to all candidate historical records of environmental parameter values ​​is calculated to obtain the mean Pearson correlation coefficient. This step is intended to provide a criterion for subsequent screening of historical records that are closer to the target monitoring period data.

[0136] Counting the number of candidate historical record environmental parameter values ​​corresponding to each type of sensor;

[0137] The number of candidate historical record environmental parameter values ​​refers to the number of candidate historical record environmental parameter values ​​of a certain type of sensor.

[0138] In the embodiment of the present disclosure, the candidate historical record environment parameter values ​​of each type of sensor are counted to obtain the number of candidate historical record environment parameter values. This step is intended to evaluate the richness of the candidate historical records.

[0139] When the number of candidate historical record environment parameter values ​​corresponding to any type of sensor is greater than or equal to a preset number threshold of the corresponding type of sensor, the candidate historical record environment parameter value whose Pearson correlation coefficient in the type of sensor is greater than or equal to the corresponding mean is determined as the target historical environment data of the type of sensor.

[0140] In the disclosed embodiment, for each type of sensor, it is first checked whether the number of candidate historical record environmental parameter values ​​is greater than or equal to a preset number threshold. If the condition is met, the candidate historical record environmental parameter values ​​whose Pearson correlation coefficient is greater than or equal to the corresponding mean are further screened out as the target historical environmental data of this type of sensor. This step is intended to ensure that the selected historical records not only have a high linear correlation with the current monitoring period data, but are also abundant in quantity to provide a reliable reference basis.

[0141] In summary, by screening candidate historical record environmental parameter values, calculating the mean of the Pearson correlation coefficient, counting the number of candidate historical records, and determining the target historical environmental data, this technology can achieve accurate screening of cave microenvironment historical data and provide strong support for subsequent data analysis and early warning. These steps together constitute a complete historical data screening process based on the Pearson correlation coefficient, which helps to improve the accuracy and reliability of cave microenvironment monitoring.

[0142] In a possible implementation, the monitoring terminal is further used for:

[0143] Constructing a collection interval duration cost matrix based on energy consumption cost and accuracy cost according to a preset collection period set corresponding to the target sensor, an environmental parameter value of the target sensor at the current collection time, and the predicted environmental parameter value;

[0144] Among them, the collection interval cost matrix: a two-dimensional matrix used to represent the total cost of the target sensor due to energy cost and accuracy cost under different collection intervals. The rows of the matrix represent different collection intervals, and the columns represent different cost factors (such as energy cost and accuracy cost). Energy cost: refers to the cost of electrical energy or other energy consumed by the sensor when collecting data. Accuracy cost: refers to the cost caused by the decrease in environmental parameter measurement accuracy due to the increase in the collection interval, which is usually related to the accuracy and reliability of the data.

[0145] In the embodiment of the present disclosure, a collection interval duration cost matrix based on energy consumption cost and accuracy cost is constructed according to a preset collection cycle set of the target sensor (i.e., a set of different collection interval durations that the sensor can adopt), as well as the environmental parameter values ​​and predicted environmental parameter values ​​at the current collection moment.

[0146] The energy cost can be calculated based on the power consumption of the sensor and the acquisition interval. For example, the higher the power consumption and the shorter the acquisition interval, the higher the energy cost. The accuracy cost can be estimated by comparing the accuracy of the predicted environmental parameter values ​​at different acquisition intervals. The longer the acquisition interval, the more changes in environmental parameters may be missed, resulting in decreased accuracy and increased accuracy cost. Each element in the cost matrix represents the sum of the energy cost and accuracy cost at a specific acquisition interval.

[0147] According to the adjustment cost matrix, multiple collection interval durations of the target sensor in the future time period starting from the current collection moment are determined, and according to the multiple collection interval durations corresponding to the multiple target sensors, the target collection interval durations of the multiple target sensors within the second preset duration are determined.

[0148] In the disclosed embodiment, after constructing the acquisition interval duration cost matrix, the cost is adjusted according to actual needs to form an adjusted cost matrix. The adjustment may involve the weighting of energy consumption cost and accuracy cost, or the correction of the cost of a specific acquisition interval duration. Next, the adjustment cost matrix is ​​used to determine the duration of multiple acquisition intervals of the target sensor in the future time period starting from the current acquisition moment. This usually involves an optimization process of minimizing the cost or maximizing the benefit in the matrix.

[0149] Finally, according to the multiple collection intervals corresponding to the multiple target sensors, the target collection interval within the second preset time (such as one day, one week, etc.) is determined. This may require comprehensive consideration of the collection requirements, energy consumption constraints, accuracy requirements, and other factors of all target sensors to find a globally optimal collection interval solution.

[0150] In summary, by constructing the collection interval time cost matrix and determining the collection interval time and other technical means, this technology can realize the intelligent optimization of the sensor collection strategy to balance the energy cost and accuracy cost, and improve the efficiency and accuracy of data collection. These steps together constitute a complete, cost-based collection interval time optimization process, which helps to improve the overall performance and sustainability of the cave microenvironment monitoring system.

[0151] In a possible implementation, the sensors for collecting different types of environmental parameters include multiple ones of the following: a temperature sensor, a relative humidity sensor, a gas sensor, an air pressure sensor, a positive and negative ion sensor, and a light sensor.

[0152] The present disclosure also provides a cave microenvironment intelligent early warning method, which is applied to a monitoring terminal in the cave microenvironment intelligent early warning system described in any of the above embodiments, such as Figure 1 As shown, the cave microenvironment intelligent early warning system includes a collection terminal and an alarm terminal that are communicatively connected to the monitoring terminal, such as Figure 2 As shown, the method includes:

[0153] In step S21, the environmental parameter values ​​of the corresponding environmental parameters in the cave microenvironment are collected by the multiple different types of sensors sent by the collection terminal in time sequence, wherein the collection terminal is placed in the cave microenvironment, and the environmental parameter values ​​of the corresponding environmental parameters in the cave microenvironment are collected by the multiple different types of sensors;

[0154] In step S22, the environmental parameter values ​​collected by the sensors of each type at each collection moment in a historical time period of a first preset length of time in reverse order from the current collection moment are taken as target environmental parameter values;

[0155] In step S23, based on the target environmental parameter value corresponding to the sensor of each type at the current collection moment and a plurality of historical record environmental parameter values ​​pre-stored by the sensor of the same type, the predicted environmental parameter value of the environmental parameter corresponding to the sensor of the type within a future time period of a second preset time length from the current collection moment is predicted;

[0156] In step S23, a plurality of sensors whose predicted environmental parameter values ​​exceed the preset parameter value threshold of the corresponding type are taken as target sensors, and the alarm type information for the cave microenvironment is determined according to the type corresponding to the target sensor;

[0157] In step S24, the alarm status information for the cave microenvironment is determined based on the predicted environmental parameter values ​​corresponding to the multiple target sensors, and based on the alarm type information and the alarm status information, an early warning instruction is generated and sent to the alarm terminal. The early warning instruction is used by the alarm terminal to issue an early warning based on the alarm type information and the alarm status information carried in the early warning instruction.

[0158] In a possible implementation, in step S23, predicting the predicted environmental parameter value of the environmental parameter corresponding to the sensor of each type at the current collection moment and a plurality of historical record environmental parameter values ​​pre-stored by the sensor of the same type within a future time period of a second preset time length from the current collection moment includes:

[0159] In step S231, according to the target environmental parameter value corresponding to each type of the sensor at the current acquisition moment and the multiple historical record environmental parameter values ​​pre-stored by the sensor of the same type, the Pearson correlation coefficient between the target environmental parameter of each type and the multiple historical record environmental parameter values ​​pre-stored by the sensor of the same type is calculated;

[0160] In step S232, according to the Pearson correlation coefficient corresponding to the target environmental parameter value of each type, target historical environmental data is determined from a plurality of historical record environmental parameter values ​​corresponding to the type;

[0161] In step S233, when the warning result information associated with the target historical environmental data corresponding to any type of the sensor represents that an alarm event occurs within a future time period of the second preset time length in sequence, the predicted environmental parameter value of the environmental parameter corresponding to the sensor of this type in the future time period of the second preset time length starting from the current collection moment is predicted based on the target environmental parameter value collected by the sensor of this type at each collection moment in the historical time period of the first preset time length.

[0162] In a possible implementation, in step S233, based on the target environmental parameter value collected by the sensor of the type at each collection moment in the historical time period of the first preset time length, predicting the predicted environmental parameter value of the environmental parameter corresponding to the sensor of the type in the future time period of the second preset time length starting from the current collection moment includes:

[0163] In step S2331, according to the target environmental parameter values ​​collected by the sensor of this type at each collection moment in the historical time period of the first preset time period, the target environmental parameter values ​​in the historical time period of the first preset time period are divided into multiple sets corresponding to the first random number through multiple random time periods of a first random number, wherein each set obtained by the first random number includes a second random number of target environmental parameter values ​​at continuous collection times of any of the random time periods;

[0164] In step S2332, according to the order in which the sets appear in the historical time period of the first preset time length, the target environment parameter values ​​included in the previous set are traversed, and the predicted parameter values ​​of the subsequent set are predicted by the deep learning model, and the model parameters of the deep learning model are adjusted according to the error between the target environment parameter value of the subsequent set and the predicted parameter value predicted by the previous set, wherein the target environment parameter value corresponding to each previous set is input into the deep learning model after the model parameters of the previous set are adjusted as the subsequent set;

[0165] In step S2333, after the model parameters are adjusted for the last set in the historical time period of the first preset duration as the subsequent set, the target environmental parameter values ​​collected by the sensor of this type at each collection moment in the historical time period of the first preset duration are input into the deep learning model after the final model parameter adjustment to obtain the predicted parameter values ​​output by the deep learning model, and to obtain the predicted environmental parameter values ​​of the environmental parameters corresponding to the sensor of this type in the future time period of the second preset duration starting from the current collection moment.

[0166] In a possible implementation, in step S2331, the target environment parameter value collected by the sensor of the type at each collection moment in the historical time period of the first preset time period is divided into multiple sets corresponding to the first random number by multiple random time periods of a first random number, including:

[0167] For any of the types of the sensors, randomly determine a first random number of the sets and randomly determine a second random number of the target environmental parameter values ​​included in each of the sets;

[0168] Constructing a plurality of the sets according to the first random number, and adding an index to each of the sets in sequence according to time;

[0169] According to the index of each of the sets and the second random number of the target environmental parameter values ​​included in each of the sets, the target environmental parameter values ​​within the historical time period of the first preset time length are divided into each of the sets, thereby completing the division of the target environmental parameter values ​​within the historical time period of the first preset time length into multiple sets corresponding to the first random number.

[0170] In a possible implementation, in step S231, the Pearson correlation coefficient between the target environmental parameter of each type and the multiple historical record environmental parameter values ​​pre-stored by the sensor of the same type is calculated according to the target environmental parameter value corresponding to the sensor of each type at the current acquisition time and the multiple historical record environmental parameter values ​​pre-stored by the sensor of the same type, including:

[0171] Calculate the first parameter value mean of the target environmental parameter value of each type of sensor according to the parameter value corresponding to each collection moment in the target environmental parameter value corresponding to each type of sensor, and calculate the second parameter value mean corresponding to each of the historical record environmental parameter values ​​pre-stored by the same type of sensor according to the parameter value corresponding to each collection moment in a plurality of historical record environmental parameter values ​​pre-stored by the same type of sensor;

[0172] Calculate, according to the mean value of the first parameter values ​​corresponding to each type of sensor and the parameter values ​​corresponding to each collection moment in the target environment parameter values ​​corresponding to the corresponding type of sensor, a first deviation between each parameter value in the target environment parameter values ​​of each type of sensor and the mean value of the first parameter values;

[0173] Calculate, according to the average of the second parameter values ​​corresponding to each of the historical record environmental parameter values ​​corresponding to each of the types of sensors and the parameter values ​​corresponding to each of the collection moments in the historical record environmental parameter values, a second deviation between each of the parameter values ​​in the historical record environmental parameter values ​​and the average of the second parameter values;

[0174] According to the first deviation and the second deviation corresponding to each type of sensor, after calculating the product of the first deviation and the second deviation corresponding to the type of sensor, summing the products to obtain a deviation product sum corresponding to the type of sensor;

[0175] Calculating the sum of squares of the first deviations corresponding to each type of sensor to obtain a first sum of squares of the deviations, and calculating the sum of squares of the second deviations corresponding to each type of sensor to obtain a second sum of squares of the deviations;

[0176] The Pearson correlation coefficient between the target environmental parameter value of each type of sensor and each of the historical record environmental parameter values ​​pre-stored by the same type of sensor is calculated based on the deviation product sum, the first deviation square sum and the second deviation square sum corresponding to each type of sensor.

[0177] In a possible implementation, in step S232, determining the target historical environment data from the plurality of historical record environment parameter values ​​corresponding to the type according to the Pearson correlation coefficient corresponding to the target environment parameter value of each type includes:

[0178] The historical record environment parameter value whose Pearson correlation coefficient of each type of sensor is greater than or equal to the preset coefficient threshold corresponding to the corresponding type of sensor is used as a candidate historical record environment parameter value;

[0179] Calculate the mean value of the Pearson correlation coefficient corresponding to the candidate historical record environmental parameter value corresponding to each type of sensor;

[0180] Counting the number of candidate historical record environmental parameter values ​​corresponding to each type of sensor;

[0181] When the number of candidate historical record environment parameter values ​​corresponding to any type of sensor is greater than or equal to a preset number threshold of the corresponding type of sensor, the candidate historical record environment parameter value whose Pearson correlation coefficient in the type of sensor is greater than or equal to the corresponding mean is determined as the target historical environment data of the type of sensor.

[0182] In a possible implementation, the monitoring terminal is further used for:

[0183] Constructing a collection interval duration cost matrix based on energy consumption cost and accuracy cost according to a preset collection period set corresponding to the target sensor, an environmental parameter value of the target sensor at the current collection time, and the predicted environmental parameter value;

[0184] According to the adjustment cost matrix, multiple collection interval durations of the target sensor in the future time period starting from the current collection moment are determined, and according to the multiple collection interval durations corresponding to the multiple target sensors, the target collection interval durations of the multiple target sensors within the second preset duration are determined.

[0185] In a possible implementation, the sensors for collecting different types of environmental parameters include multiple ones of the following: a temperature sensor, a relative humidity sensor, a gas sensor, an air pressure sensor, a positive and negative ion sensor, and a light sensor.

[0186] The method steps involved in the monitoring terminal in the above method embodiment are all described in detail in the system in the above embodiment, and will not be repeated here.

[0187] Figure 3 The monitoring terminal 100 shown includes: a processor 1001 and a memory 1003. The processor 1001 and the memory 1003 are connected, such as through a bus 1002. Optionally, the cave microenvironment intelligent early warning method 100 may also include a communication component 1004, which can be used for data interaction between the device 100 and other devices, such as data transmission and / or data reception. It should be noted that in actual scheduling, the communication component 1004 is not limited to one, and the structure of the monitoring terminal 100 does not constitute a limitation on the embodiments of the present application.

[0188] Processor 1001 may be a CPU (Central Processing Unit), a general purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of this application. Processor 1001 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0189] The bus 1002 may include a path to transmit information between the above components. The bus 1002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus 1002 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0190] The memory 1003 can be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compressed optical disk, laser disk, optical disk, digital versatile disk, Blu-ray disk, etc.), magnetic disk storage medium, other magnetic storage devices, or any other medium that can be used to carry or store program code and can be read by a computer, without limitation herein.

[0191] The memory 1003 is used to store program codes for executing the embodiments of the present disclosure, and the execution is controlled by the processor 1001. The processor 1001 is used to execute the program codes stored in the memory 1003 to implement the steps shown in the embodiment of the cave microenvironment intelligent early warning method.

[0192] The embodiment of the present disclosure also provides a computer-readable storage medium having program code stored thereon. When the program code is executed by a processor, the steps and corresponding contents of the aforementioned cave microenvironment intelligent early warning method embodiment can be implemented.

[0193] The preferred embodiments of the present disclosure are described in detail above in conjunction with the accompanying drawings; however, the present disclosure is not limited to the specific details in the above embodiments; within the technical concept of the present disclosure, various changes, modifications, substitutions and variations may be made to these embodiments, and these changes, modifications, substitutions and variations all fall within the protection scope of the present disclosure.

[0194] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any appropriate manner without contradiction, and they should also be regarded as the contents disclosed in this disclosure. In order to avoid unnecessary repetition, this disclosure will not further describe various possible combinations. The technical scope of this application is not limited to the contents in the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. An intelligent early warning system for cave microenvironment, characterized in that: The system comprises: A collection terminal, a monitoring terminal connected to the collection terminal in communication, and an alarm terminal connected to the monitoring terminal in communication; The acquisition terminal is configured with a plurality of sensors for collecting different types of environmental parameters. The acquisition terminal is used to be placed in the cave microenvironment, and to collect environmental parameter values ​​of corresponding environmental parameters in the cave microenvironment through a plurality of different types of sensors, and to send the environmental parameter values ​​collected by the different types of sensors to the monitoring terminal in chronological order; The monitoring terminal is used to use the environmental parameter values ​​collected by the sensors of each type at each collection moment in a historical time period of a first preset time length in reverse order from the current collection moment as target environmental parameter values; According to the target environmental parameter value corresponding to the sensor of each type at the current collection moment and a plurality of historical record environmental parameter values ​​pre-stored by the sensor of the same type, predict the predicted environmental parameter value of the environmental parameter corresponding to the sensor of the type within a future time period of a second preset time length starting from the current collection moment; Taking multiple sensors whose predicted environmental parameter values ​​exceed the preset parameter value threshold of the corresponding type as target sensors, determining the alarm type information for the cave microenvironment according to the type corresponding to the target sensor; determining the alarm status information for the cave microenvironment according to the predicted environmental parameter values ​​corresponding to the multiple target sensors, and generating and sending an early warning instruction to the alarm terminal according to the alarm type information and the alarm status information; The alarm terminal is used to receive the early warning instruction sent by the monitoring terminal, and issue an early warning according to the alarm type information and alarm status information carried in the early warning instruction.

2. The system according to claim 1, characterized in that The monitoring terminal is specifically used for: Calculate the Pearson correlation coefficient between the target environmental parameter of each type and the multiple historical record environmental parameter values ​​pre-stored by the sensor of the same type according to the target environmental parameter value corresponding to the sensor of each type at the current acquisition moment and the multiple historical record environmental parameter values ​​pre-stored by the sensor of the same type; Determine target historical environmental data from a plurality of historical record environmental parameter values ​​corresponding to the type according to the Pearson correlation coefficient corresponding to the target environmental parameter value of each type; When the warning result information associated with the target historical environmental data corresponding to any type of sensor represents that an alarm event occurs within a future time period of the second preset time length in sequence, the predicted environmental parameter value of the environmental parameter corresponding to the sensor of this type in the future time period of the second preset time length starting from the current collection moment is predicted based on the target environmental parameter value collected by the sensor of this type at each collection moment in the historical time period of the first preset time length.

3. The system according to claim 2, characterized in that The monitoring terminal is specifically used for: According to the target environmental parameter values ​​collected by the sensor of this type at each collection moment in the historical time period of the first preset time period, the target environmental parameter values ​​in the historical time period of the first preset time period are divided into multiple sets corresponding to the first random number through multiple random time periods of a first random number, wherein each set obtained by the first random number includes a second random number of target environmental parameter values ​​at continuous collection times of any of the random time periods; According to the order in which the sets appear in the historical time period of the first preset time length, the target environment parameter values ​​included in the previous set are traversed, and the predicted parameter values ​​of the subsequent set are predicted by the deep learning model, and the model parameters of the deep learning model are adjusted according to the error between the target environment parameter value of the subsequent set and the predicted parameter value predicted by the previous set, wherein the target environment parameter value corresponding to each previous set is input into the deep learning model after the model parameter of the previous set is adjusted as the subsequent set; After the set that is last in the historical time period of the first preset duration is used as the subsequent set for model parameter adjustment, the target environmental parameter values ​​collected by the sensor of this type at each collection moment in the historical time period of the first preset duration are input into the deep learning model after the final model parameter adjustment to obtain the predicted parameter values ​​output by the deep learning model, and to obtain the predicted environmental parameter values ​​of the environmental parameters corresponding to the sensor of this type in the future time period of the second preset duration starting from the current collection moment.

4. The system according to claim 3, characterized in that The monitoring terminal is specifically used for: For any of the types of the sensors, randomly determine a first random number of the sets and randomly determine a second random number of the target environmental parameter values ​​included in each of the sets; Constructing a plurality of the sets according to the first random number, and adding an index to each of the sets in sequence according to time; According to the index of each of the sets and the second random number of the target environmental parameter values ​​included in each of the sets, the target environmental parameter values ​​within the historical time period of the first preset time length are divided into each of the sets, thereby completing the division of the target environmental parameter values ​​within the historical time period of the first preset time length into multiple sets corresponding to the first random number.

5. The system according to claim 2, characterized in that The monitoring terminal is specifically used for: Calculate the first parameter value mean of the target environmental parameter value of each type of sensor according to the parameter value corresponding to each collection moment in the target environmental parameter value corresponding to each type of sensor, and calculate the second parameter value mean corresponding to each of the historical record environmental parameter values ​​pre-stored by the same type of sensor according to the parameter value corresponding to each collection moment in a plurality of historical record environmental parameter values ​​pre-stored by the same type of sensor; Calculate, according to the mean value of the first parameter values ​​corresponding to each type of sensor and the parameter values ​​corresponding to each collection moment in the target environment parameter values ​​corresponding to the corresponding type of sensor, a first deviation between each parameter value in the target environment parameter values ​​of each type of sensor and the mean value of the first parameter values; Calculate, according to the average of the second parameter values ​​corresponding to each of the historical record environmental parameter values ​​corresponding to each of the types of sensors and the parameter values ​​corresponding to each of the collection moments in the historical record environmental parameter values, a second deviation between each of the parameter values ​​in the historical record environmental parameter values ​​and the average of the second parameter values; According to the first deviation and the second deviation corresponding to each type of sensor, after calculating the product of the first deviation and the second deviation corresponding to the type of sensor, summing the products to obtain a deviation product sum corresponding to the type of sensor; Calculating the sum of squares of the first deviations corresponding to each type of sensor to obtain a first sum of squares of the deviations, and calculating the sum of squares of the second deviations corresponding to each type of sensor to obtain a second sum of squares of the deviations; The Pearson correlation coefficient between the target environmental parameter value of each type of sensor and each of the historical record environmental parameter values ​​pre-stored by the same type of sensor is calculated based on the deviation product sum, the first deviation square sum and the second deviation square sum corresponding to each type of sensor.

6. The system according to claim 5, characterized in that The monitoring terminal is specifically used for: The historical record environment parameter value whose Pearson correlation coefficient of each type of sensor is greater than or equal to the preset coefficient threshold corresponding to the corresponding type of sensor is used as a candidate historical record environment parameter value; Calculate the mean value of the Pearson correlation coefficient corresponding to the candidate historical record environmental parameter value corresponding to each type of sensor; Counting the number of candidate historical record environmental parameter values ​​corresponding to each type of sensor; When the number of candidate historical record environment parameter values ​​corresponding to any type of sensor is greater than or equal to a preset number threshold of the corresponding type of sensor, the candidate historical record environment parameter value whose Pearson correlation coefficient in the type of sensor is greater than or equal to the corresponding mean is determined as the target historical environment data of the type of sensor.

7. The system according to claim 1, characterized in that The monitoring terminal is also used for: Constructing a collection interval duration cost matrix based on energy consumption cost and accuracy cost according to a preset collection period set corresponding to the target sensor, an environmental parameter value of the target sensor at the current collection time, and the predicted environmental parameter value; According to the adjustment cost matrix, multiple collection interval durations of the target sensor in the future time period starting from the current collection moment are determined, and according to the multiple collection interval durations corresponding to the multiple target sensors, the target collection interval durations of the multiple target sensors within the second preset duration are determined.

8. The system according to any one of claims 1 to 7, characterized in that: The sensors for collecting different types of environmental parameters include multiple ones of the following: temperature sensor, relative humidity sensor, gas sensor, air pressure sensor, positive and negative ion sensor, and light sensor.

9. A cave microenvironment intelligent early warning method, characterized in that: A monitoring terminal used in the cave microenvironment intelligent early warning system according to any one of claims 1 to 8, wherein the cave microenvironment intelligent early warning system comprises a collection terminal and an alarm terminal that are communicatively connected to the monitoring terminal, and the method comprises: Receiving the environmental parameter values ​​of the corresponding environmental parameters in the cave microenvironment collected by the multiple different types of sensors sent by the collection terminal in time sequence, wherein the collection terminal is placed in the cave microenvironment, and collects the environmental parameter values ​​of the corresponding environmental parameters in the cave microenvironment through the multiple different types of sensors; The environmental parameter values ​​collected by the sensors of each type at each collection moment in a historical time period of a first preset length of time in reverse order from the current collection moment are used as target environmental parameter values; According to the target environmental parameter value corresponding to the sensor of each type at the current collection moment and a plurality of historical record environmental parameter values ​​pre-stored by the sensor of the same type, predict the predicted environmental parameter value of the environmental parameter corresponding to the sensor of the type within a future time period of a second preset time length starting from the current collection moment; Taking multiple sensors whose predicted environmental parameter values ​​exceed the preset parameter value thresholds of corresponding types as target sensors, and determining the alarm type information for the cave microenvironment according to the types corresponding to the target sensors; According to the predicted environmental parameter values ​​corresponding to the multiple target sensors, the alarm status information for the cave microenvironment is determined, and according to the alarm type information and the alarm status information, an early warning instruction is generated and sent to the alarm terminal, and the early warning instruction is used for the alarm terminal to issue an early warning according to the alarm type information and the alarm status information carried in the early warning instruction.

10. The method according to claim 9, characterized in that The predicting, based on the target environmental parameter value corresponding to the sensor of each type at the current collection moment and a plurality of historical record environmental parameter values ​​pre-stored by the sensor of the same type, of the predicted environmental parameter value of the environmental parameter corresponding to the sensor of the type within a future time period of a second preset time length starting from the current collection moment, includes: Calculate the Pearson correlation coefficient between the target environmental parameter of each type and the multiple historical record environmental parameter values ​​pre-stored by the sensor of the same type according to the target environmental parameter value corresponding to the sensor of each type at the current acquisition moment and the multiple historical record environmental parameter values ​​pre-stored by the sensor of the same type; Determine target historical environmental data from a plurality of historical record environmental parameter values ​​corresponding to the type according to the Pearson correlation coefficient corresponding to the target environmental parameter value of each type; When the warning result information associated with the target historical environmental data corresponding to any type of sensor represents that an alarm event occurs within a future time period of the second preset time length in sequence, the predicted environmental parameter value of the environmental parameter corresponding to the sensor of this type in the future time period of the second preset time length starting from the current collection moment is predicted based on the target environmental parameter value collected by the sensor of this type at each collection moment in the historical time period of the first preset time length.

Images