Environment detection and inspection robot and recognition method of environment detection and inspection robot

Abstract

The invention provides an environment detection inspection robot and an identification method of the environment detection inspection robot. The environment detection inspection robot comprises a sensing unit which comprises a plurality of sensors and is used for detecting environment information in an explosive warehouse, and the environment information at least comprises one or more of gas concentration, temperature, humidity and particulate matter concentration; the walking unit is used for realizing the movement of the environment detection inspection robot in the explosive warehouse; the controller is in communication connection with the sensing unit and the walking unit, the controller is used for controlling the walking unit to work, and the controller is further used for determining whether the environment in the explosive warehouse is abnormal or not according to the environment information. According to the scheme, the problem that in the prior art, most of robots for inspecting explosive warehouses are provided with single temperature sensors, environment detection is achieved through temperature monitoring, and the accuracy of recognition results is low is solved.

Description

Technical Field

[0001] The present application relates to the technical field of environmental monitoring, and in particular to an environmental detection inspection robot, an identification method for an environmental detection inspection robot, an identification device for an environmental detection inspection robot, and an environmental detection system. Background Art

[0002] Explosives are highly dangerous. Once a fire, explosion or leakage occurs, it may cause serious casualties and property losses. Through environmental testing, changes in the environment inside and around explosive warehouses can be monitored in a timely manner, potential safety hazards can be discovered early, and the safety of warehouses and surrounding environments can be guaranteed. Many countries and regions have formulated strict laws and regulations on environmental monitoring of explosive warehouses, requiring warehouse managers to conduct regular monitoring and testing to ensure the safe storage and transportation of explosives. Environmental testing helps ensure that explosive warehouses comply with the requirements of laws and regulations and avoid legal liability and penalties for illegal acts. It can also help identify potential risk factors inside and around warehouses, such as gas leaks, abnormal temperatures, fire hazards, etc., which helps prevent accidents. By timely discovering and handling environmental anomalies, the probability of accidents in explosive warehouses can be effectively reduced, ensuring the safety of personnel and property. However, most of the current robots for patrolling explosive warehouses are equipped with a single temperature sensor to achieve environmental testing through temperature monitoring, and the accuracy of such identification results is low. Summary of the invention

[0003] The main purpose of the present application is to provide an environmental detection inspection robot, an identification method for an environmental detection inspection robot, an identification device for an environmental detection inspection robot and an environmental detection system, so as to at least solve the problem that most of the robots for inspecting explosive warehouses in the prior art are equipped with a single temperature sensor and realize environmental detection through temperature monitoring, and the accuracy of such identification results is low.

[0004] In order to achieve the above-mentioned purpose, according to one aspect of the present application, there is provided an environment detection inspection robot, comprising: a sensing unit, comprising a plurality of sensors, the sensing unit being used to detect environmental information within an explosive warehouse, wherein the environmental information comprises at least one or more of gas concentration, temperature, humidity and particle concentration; a walking unit being used to enable the movement of the environment detection inspection robot within the explosive warehouse; and a controller being respectively communicatively connected to the sensing unit and the walking unit, the controller being used to control the operation of the walking unit, and the controller being further used to determine whether the environment within the explosive warehouse is abnormal based on the environmental information.

[0005] Optionally, the sensing unit includes: a gas sensor for detecting the gas concentration in the explosive warehouse; a temperature sensor for detecting the temperature in the explosive warehouse; a humidity sensor for detecting the humidity in the explosive warehouse; and a particulate matter sensor for detecting the particulate matter concentration in the explosive warehouse.

[0006] Optionally, the environmental detection inspection robot further includes: an acoustic radar, which is communicatively connected to the controller, and the acoustic radar is used to detect sound signals in the explosive warehouse.

[0007] According to another aspect of the present application, there is provided an identification method based on any one of the environmental detection inspection robots, the method further comprising: acquiring the environmental information, wherein the environmental information comprises one or more of gas concentration, temperature, humidity and particle concentration; fusing a plurality of the environmental information to obtain fused environmental information, wherein the fusion method comprises one or more of weighted average method, Kalman filtering method, multi-Bayesian estimation method, DS evidence reasoning method, production rules and fuzzy logic reasoning; constructing an environmental detection model, wherein the environmental detection model is obtained by training with a CNN algorithm using multiple sets of training data, each set of training data in the multiple sets of training data comprises historical fused environmental information obtained within a historical time period, and a historical first detection result corresponding to the historical fused environmental information, wherein the historical first detection result is used to characterize whether the historical environment in the explosive warehouse is abnormal; inputting the fused environmental information into the environmental detection model to obtain a first detection result corresponding to the fused environmental information.

[0008] Optionally, obtaining the environmental information includes: obtaining initial environmental information; performing mean filtering on the initial environmental information to obtain first environmental information; normalizing the first environmental information to obtain second environmental information; and performing feature extraction on the second environmental information to obtain the environmental information.

[0009] Optionally, the environmental detection inspection robot also includes an acoustic radar, which is communicatively connected to the controller, and the acoustic radar is used to detect sound signals in the explosive warehouse. After obtaining the environmental information, the method also includes: obtaining the sound signal; performing voiceprint recognition on the sound signal to determine whether there is an explosion sound.

[0010] Optionally, voiceprint recognition is performed on the sound signal to determine whether there is an explosion sound, including: performing Fourier transform on the sound signal to obtain a spectrum of the sound signal; taking the logarithm of the energy signal of the spectrum to obtain a sound feature graph; constructing a sound detection model, wherein the sound detection model is trained using a convolution algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes a historical sound feature graph obtained within a historical time period, and a historical second detection result corresponding to the historical sound feature graph, wherein the historical second detection result is used to characterize whether there is a historical explosion sound in the explosive warehouse; the sound feature graph is input into the sound detection model to obtain a second detection result corresponding to the sound feature graph.

[0011] Optionally, after performing voiceprint recognition on the sound signal to determine whether there is an explosion sound, the method further includes: when it is determined that there is the explosion sound, determining the direction of the sound source based on sound detection technology, wherein the sound detection technology includes one or more of a microphone array method, an ultrasonic positioning method and a spectrum analysis method.

[0012] According to another aspect of the present application, there is provided an identification device based on any one of the environmental detection inspection robots, the device further comprising: a first acquisition unit, for acquiring the environmental information, wherein the environmental information comprises one or more of gas concentration, temperature, humidity and particle concentration; a fusion unit, for fusing a plurality of the environmental information to obtain fused environmental information, wherein the fusion method comprises one or more of weighted average method, Kalman filtering method, multi-Bayesian estimation method, DS evidence reasoning method, production rules and fuzzy logic reasoning; a construction unit, for constructing an environmental detection model, wherein the environmental detection model is obtained by training with a CNN algorithm using multiple sets of training data, each set of training data in the multiple sets of training data comprises historical fused environmental information obtained within a historical time period, and a historical first detection result corresponding to the historical fused environmental information, wherein the historical first detection result is used to characterize whether the historical environment in the explosive warehouse is abnormal; a processing unit, for inputting the fused environmental information into the environmental detection model to obtain a first detection result corresponding to the fused environmental information.

[0013] According to another aspect of the present application, an environmental detection system is provided, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include an identification method for executing any one of the environmental detection inspection robots.

[0014] Applying the technical solution of the present application, a new inspection robot for use in explosive warehouses is designed. By installing multiple sensors on the inspection robot, the inspection robot collects information on multiple environments of the explosive warehouse while moving in the explosive warehouse, thereby avoiding the problem of low accuracy of single temperature detection. Environmental detection is achieved by collecting environmental information through multiple sensors, ensuring a high accuracy of the recognition results. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The drawings constituting part of the present application are used to provide a further understanding of the present application. The exemplary embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0016] Figure 1 The schematic diagram of the structure of the environmental detection inspection robot of this scheme is shown;

[0017] Figure 2 A schematic diagram showing the side view of the environmental detection inspection robot of the present solution;

[0018] Figure 3 A hardware structure block diagram of a mobile terminal for performing an identification method of an environment detection inspection robot provided in an embodiment of the present application is shown;

[0019] Figure 4 A schematic diagram of a process flow of an environmental detection inspection robot according to an embodiment of the present application is shown;

[0020] Figure 5 A schematic diagram showing the main functions of the solution;

[0021] Figure 6 A schematic diagram showing the basic principles of sensor data fusion and cross-validation of the present scheme;

[0022] Figure 7 A structural block diagram of an identification device of an environment detection inspection robot provided according to an embodiment of the present application is shown.

[0023] The above drawings include the following reference numerals:

[0024] 102. Processor; 104. Memory; 106. Transmission device; 108. Input / output device; 10. Body; 11. Travel unit; 12. Driving wheel; 13. Driven wheel; 14. Track; 15. Servo motor; 16. Turntable; 17. Gas sensor; 18. Temperature sensor; 19. Humidity sensor; 20. Particle sensor; 21. Infrared sensor; 22. Acoustic radar; 23. Panoramic camera; 24. Detection port. DETAILED DESCRIPTION

[0025] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0026] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0027] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0028] At present, most robots that inspect explosive warehouses cannot fully detect environmental data, and there is no unified processing interface when the detection data is transmitted to the back end. This makes it impossible to respond to some data in a timely manner. For example, for temperature data, general inspection robots can only record the temperature data of the current task, or simply make judgments based on temperature requirements, and issue warnings when temperature data exceeds the standard. When the temperature data is abnormal, it is unable to collect and summarize other data in a timely manner and send them to the remote control center for new comprehensive judgments.

[0029] As introduced in the background technology, most of the robots used for patrolling explosive warehouses in the prior art are equipped with a single temperature sensor to achieve environmental detection through temperature monitoring. The accuracy of such identification results is low. To solve the above problems, the embodiments of the present application provide an environmental detection patrol robot, an identification method for an environmental detection patrol robot, an identification device for an environmental detection patrol robot, and an environmental detection system.

[0030] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0031] The present application also provides an environmental detection inspection robot, comprising:

[0032] A sensing unit, comprising a plurality of sensors, wherein the sensing unit is used to detect environmental information in the explosive warehouse, wherein the environmental information at least includes one or more of gas concentration, temperature, humidity and particle concentration;

[0033] A walking unit, used to enable the environmental detection inspection robot to move within the above-mentioned explosive warehouse;

[0034] The controller is respectively connected to the sensing unit and the walking unit for communication. The controller is used to control the walking unit to work. The controller is also used to determine whether the environment in the explosive warehouse is abnormal based on the environmental information.

[0035] Through this embodiment, a new inspection robot for use in explosive warehouses is designed. By installing multiple sensors on the inspection robot, the inspection robot collects information on multiple environments of the explosive warehouse while moving in the explosive warehouse, thereby avoiding the problem of low accuracy of single temperature detection. Environmental detection is achieved by collecting environmental information through multiple sensors, ensuring a high accuracy of the recognition results.

[0036] Specifically, this solution aims to provide an environmental detection inspection robot, which can make linkage judgments based on multiple different detection results during the inspection process, integrate various detection results, uniformly generate a detection report and send it to a remote control center.

[0037] Specifically, Figure 1 and Figure 2 As shown, the environment detection inspection robot includes a body 10 and a walking unit 11, and the walking unit is a crawler walking mechanism, including a wheel frame, a driving wheel 12, a driven wheel 13 and a crawler 14, the driving wheel is driven by a servo motor 15, and the body part is connected to the walking mechanism through a turntable 16 that can rotate 360 ​​degrees, which includes a casing and a power supply, a main control board (i.e., a controller) and various sensors arranged in the casing. The driving wheel is the power source of the environment detection inspection robot, and is driven by a motor to rotate, thereby pushing the robot forward or backward. The driven wheels mainly play the role of supporting and assisting movement in the environment detection inspection robot. They are not directly driven by a motor, but passively rotate with the rotation of the driving wheel. The crawler is an important component for the environment detection inspection robot to move in complex terrain, and is particularly suitable for harsh environments such as rugged, muddy and slippery.

[0038] In the specific implementation process, Figure 1As shown, the sensing unit includes a gas sensor 17, a temperature sensor 18, a humidity sensor 19 and a particle sensor 20. The gas sensor is used to detect the gas concentration in the explosive warehouse; the temperature sensor is used to detect the temperature in the explosive warehouse; the humidity sensor is used to detect the humidity in the explosive warehouse; and the particle sensor is used to detect the particle concentration in the explosive warehouse. Of course, the sensing unit is not limited to the above sensors, but can also be any other feasible sensor. For example, a smoke sensor, a video surveillance sensor, an infrared sensor 21, and the like.

[0039] In some embodiments, such as Figure 1 As shown, the above-mentioned environmental detection inspection robot also includes an acoustic radar 22, which is communicatively connected to the above-mentioned controller, and the above-mentioned acoustic radar is used to detect sound signals in the above-mentioned explosive warehouse.

[0040] In this scheme, acoustic radar is a device that uses sound waves for detection and positioning. It emits sound waves and receives sound waves reflected from obstacles, and determines the position, shape and distance of obstacles by calculating the round-trip time and direction of the sound waves. Unlike optical radar (lidar), acoustic radar is not limited by light conditions and can work normally in dark or smoky environments. The main function of acoustic radar is to monitor and identify sound signals in explosive warehouses. By monitoring the sound signals in the warehouse, acoustic radar can detect abnormal sounds in time, such as leakage sounds, explosion precursor sounds, etc., so as to give early warning and reduce the risk of accidents. Acoustic radar is not limited by light and can work normally in dark or smoky environments, which enhances the monitoring capabilities of warehouses, ensures all-weather and all-round monitoring, reduces the frequency and labor intensity of manual inspections, and improves inspection efficiency.

[0041] Specifically, Figure 2 As shown, the environment detection inspection robot also includes a panoramic camera 23 and a turntable. The panoramic camera is used to collect images in the warehouse, and the turntable is used to support the body rotation of the environment detection inspection robot. Specifically, Figure 1 The detection port 24 shown in the figure is used to send the collected data to the internal processor or controller.

[0042] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 3 FIG. 1 is a hardware structure block diagram of a mobile terminal of an environmental detection inspection robot identification method according to an embodiment of the present invention. Figure 3 As shown, the mobile terminal may include one or more ( Figure 3Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 3 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 3 More or fewer components as shown, or with Figure 3 Different configurations are shown.

[0043] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the display method of device information in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the above method is implemented. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof. The transmission device 106 is used to receive or send data via a network. The above-mentioned specific examples of the network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0044] In this embodiment, a method for identifying an environmental detection inspection robot running on a mobile terminal, a computer terminal or a similar computing device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0045] Figure 4 It is a flow chart of an identification method of an environmental detection inspection robot according to an embodiment of the present application.

[0046] like Figure 4 As shown, the method comprises the following steps:

[0047] Step S201, obtaining the above-mentioned environmental information, wherein the above-mentioned environmental information includes one or more of gas concentration, temperature, humidity and particle concentration;

[0048] Specifically, the environmental information in the explosive warehouse can be obtained. Gas concentration refers to the content of one or more gases in the environment. Temperature is a basic physical quantity in the environment, reflecting the thermal state of the environment. Humidity refers to the content of water vapor in the air. Particle concentration refers to the number or mass of solid or liquid particles suspended in the air.

[0049] Step S202, fusing the plurality of the above environmental information to obtain fused environmental information, wherein the fusion method includes one or more of a weighted average method, a Kalman filter method, a multi-Bayesian estimation method, a DS evidence reasoning method, a production rule, and a fuzzy logic reasoning method;

[0050] Specifically, through fusion processing, multiple environmental information are integrated into more comprehensive, accurate and reliable fused environmental information.

[0051] Step S203, constructing an environment detection model, wherein the environment detection model is obtained by training with a CNN algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes historical fusion environment information obtained in a historical time period and a historical first detection result corresponding to the historical fusion environment information, wherein the historical first detection result is used to characterize whether the historical environment in the explosive warehouse is abnormal;

[0052] Specifically, an environmental detection model is constructed. This model will be used to detect the environment in the explosive warehouse to determine whether the environment is abnormal. The model is trained using a convolutional neural network (CNN) algorithm, and the training data is multiple sets of historical data. The historical first detection result refers to the result corresponding to the above-mentioned historical fusion environmental information, which is used to characterize whether the historical environment in the explosive warehouse is abnormal. The environmental detection model trained by the CNN algorithm can learn the complex characteristics of environmental information, thereby improving the accuracy of detection.

[0053] Step S204: input the above-mentioned fused environment information into the above-mentioned environment detection model to obtain a first detection result corresponding to the above-mentioned fused environment information.

[0054] Specifically, the environmental detection model can be used to detect whether the environment is abnormal. Compared with traditional solutions, it is more intelligent and has more accurate detection and identification, and introduces machine learning algorithms for anomaly analysis.

[0055] Through this embodiment, the inspection robot collects information on multiple environments of the explosive warehouse while moving in the explosive warehouse, and determines whether the environment in the explosive warehouse is abnormal through the paradigm of multi-data fusion, thereby avoiding the problem of low accuracy of single temperature detection. Environmental detection is achieved by collecting environmental information through multiple sensors, ensuring a high accuracy of the recognition results.

[0056] Specifically, in current conventional environmental detection robots, a threshold is usually set for a single data to make a judgment, and there is a lack of correlation between the various data, which often results in false alarms. The key point of the body and walking mechanism of the solution of this application, as well as the various integrated sensors, is to improve the accuracy of abnormality detection and reduce false alarms through abnormality analysis. When one sensor data exceeds the threshold range, the multi-sensor data fusion and cross-validation method is used.

[0057] Specifically, multi-sensor information fusion is an information processing process that uses computer technology to automatically analyze and synthesize information and data from multiple sensors or multiple sources under certain criteria to complete the required decision-making and estimation. The principle of multi-sensor information fusion is mainly to carry out multi-level and multi-space information complementation and optimization combination processing of various sensors, and finally produce a consistent interpretation of the observed environment. In this process, it is necessary to make full use of multi-source data for reasonable control and use, and the ultimate goal of information fusion is to derive more useful information based on the separated observation information obtained by each sensor through multi-level and multi-faceted combination of information. This not only takes advantage of the mutual cooperation of multiple sensors, but also comprehensively processes data from other information sources to improve the intelligence of the entire sensor system. This algorithm mainly includes pre-fusion algorithm and post-fusion algorithm. The pre-fusion algorithm fuses all data together at the original layer. The fused data is like a "sensor", and this "sensor" is integrated. On this "sensor", the perception algorithm is applied, and finally a perception result will be output. The post-fusion algorithm is the target data generated by each sensor independently processing. Each sensor has its own independent perception, such as laser radar has its own perception, camera has its own perception, and millimeter wave radar also has its own perception. When all sensors have completed the generation of target data, the main processor will perform data fusion. The proposal of this invention adopts the post-fusion algorithm. Currently, the common post-fusion algorithms mainly include weighted average method, Kalman filter method, multi-Bayesian estimation method, DS evidence reasoning method, production rules, fuzzy logic reasoning, etc.

[0058] In the specific implementation process, the above-mentioned environmental information can be obtained through the following steps: obtaining initial environmental information; performing mean filtering on the above-mentioned initial environmental information to obtain first environmental information; normalizing the above-mentioned first environmental information to obtain second environmental information; performing feature extraction on the above-mentioned second environmental information to obtain the above-mentioned environmental information.

[0059] In this scheme, the initial environmental information can be preprocessed, and the preprocessing includes mean filtering, normalization and feature extraction. Through mean filtering and normalization, the noise and redundant information in the original data can be removed, and the quality and reliability of the data can be improved. Normalization is to scale the data to a specific range (usually 0 to 1 or -1 to 1) to facilitate subsequent data analysis and processing. Feature extraction is the process of extracting the most representative information or pattern from the data. Feature extraction can reduce the dimension and complexity of the data and reduce the computational complexity of subsequent analysis and processing.

[0060] Specifically, Figure 5 As shown, this solution can also perform charge and discharge control, patrol control and communication. In addition to generating patrol routes, patrol control also requires real-time obstacle avoidance control. For this solution, the key lies in data analysis. Specifically, data analysis includes statistical analysis, visualization analysis, abnormal analysis and data fusion.

[0061] Specifically, the charge and discharge control is explained in two steps, charge and discharge.

[0062] Charging: The robot should be equipped with a power monitoring system to monitor the battery power in real time. When the power is lower than the preset safety threshold, the robot should send a low power alarm to the background monitoring center. After receiving the low power alarm, the background monitoring center will send a charging instruction to the robot body according to the current position of the robot and the location of the charging station. The instruction should include information such as the coordinates of the charging station and the best path. After receiving the charging instruction, the robot body uses the internal navigation system and map information to autonomously plan an optimal path from the current position to the charging station. When planning the path, the robot needs to consider factors such as obstacles and safe distances to ensure the safety and feasibility of the path. When the robot approaches the charging station, it needs to perform charging docking operations. This usually includes detecting the voltage and current of the charging electrode, the digital signal of the photoelectric switch, etc. If the detection conditions are met (that is, there is voltage and current on the electrode, and a digital signal is generated by the photoelectric detection switch), the robot should stop immediately for charging. If the detection conditions are not met, the robot needs to re-plan the path or perform troubleshooting until it is successfully docked and charged. During the charging process, the robot should continuously monitor the battery power and charging status. If any abnormality occurs (such as battery overheating, abnormal charging speed, etc.), the robot should stop charging immediately and send an alarm to the background monitoring center.

[0063] Discharge: According to the inspection requirements of the explosive warehouse and the performance parameters of the robot, a detailed inspection task plan is formulated. The task plan should include information such as inspection points, inspection routes, and inspection time. Before executing the task, the robot needs to evaluate whether the current power meets the task requirements. If the power is insufficient, it needs to be charged first. The robot performs inspection operations according to the task plan. During the inspection process, the robot needs to use sensors, cameras and other equipment to monitor the warehouse environment in real time. If an abnormal situation is found (such as abnormal temperature, gas leakage, etc.), the robot should immediately send an alarm to the background monitoring center. During the execution of the task, the robot needs to continuously monitor the power consumption. If the power is lower than the preset safety threshold, the robot should return to the charging station in advance for charging to avoid mission failure or safety hazards due to power exhaustion.

[0064] Specifically, for patrol control, before deploying the patrol robot, it is necessary to conduct a comprehensive assessment of the environment of the explosive warehouse, including the layout of the warehouse, explosion-proof level, temperature, humidity and other conditions. Ensure that the warehouse environment meets the robot's operating requirements to avoid potential safety hazards. According to the actual situation of the warehouse, select the appropriate type of patrol robot (such as explosion-proof, wheeled, etc.), and configure the corresponding charging station and communication system. Ensure that the robot communicates smoothly with the background monitoring system to transmit data and receive instructions in real time. Use the robot's built-in laser navigation technology to scan the warehouse with high precision and build a two-dimensional grid map or a three-dimensional point cloud map. Import the map into the robot's background database system to achieve autonomous path planning and precise positioning. Set the inspection route and inspection points on the map according to the inspection requirements. Consider factors such as obstacles and safety distances in the warehouse, optimize the path planning, and ensure that the robot can complete the task safely and efficiently. Set the inspection task in the background monitoring system, including the inspection route, inspection point, inspection frequency, etc. Send the task to the robot body to ensure that the robot can perform the inspection task according to the predetermined plan. After entering the warehouse, the robot will autonomously plan the navigation path according to the set tasks and map information. It will use sensors, cameras and other equipment to monitor the surrounding environment in real time to avoid collisions and safety hazards. During the inspection process, the robot needs to update its own location information in real time and compare it with the map. If it is found that the actual path deviates from the planned path, the robot needs to automatically adjust the navigation strategy to ensure that the inspection continues along the planned route. When encountering obstacles or complex environments, the robot needs to have obstacle avoidance and path replanning capabilities to ensure the smooth completion of the inspection task. During the inspection process, the robot needs to use sensors, cameras and other equipment to monitor the warehouse environment data (such as temperature, humidity, gas concentration, etc.) in real time. The collected data is uploaded to the background monitoring system in real time for analysis and alarm processing. When the robot completes the inspection task, it needs to automatically return to the charging station for charging or standby. Send a task completion notification to the background monitoring system for subsequent processing by the management personnel. The background monitoring system analyzes and processes the data uploaded by the robot. Based on the analysis results, potential safety hazards or abnormal situations are found, and corresponding treatment measures are taken.

[0065] Specifically, Figure 6 As shown in the figure, the inspection robot equipped with sensors such as gas sensors, temperature sensors, humidity sensors and particle detection sensors inspects along the inspection route and collects environmental data in real time under the control of the patrol control. The sensor data is first converted to digital by the ADC chip to convert the analog signals of each sensor into digital signals. The digital signals of the sensors are then further preprocessed, which includes:

[0066] 1) Perform mean filtering on sensor data to smooth the data and reduce the impact of random noise on the data;

[0067] 2) Perform normalization on the filtered data. The specific process is as follows:

[0068] (1) Select 5 minutes of sensor data to form a data set, traverse each value in the data set, and determine the maximum value x max and the minimum data x min ;

[0069] (2) Perform normalization according to the following formula:

[0070]

[0071] Among them, x is the original data, x min is the minimum value in the data set, x max is the maximum value in the data set, and x′ is the normalized data, with a range of [0, 1]. After normalization, the data from different sensors are scaled to the same range. This process only changes the scale of the data, but does not change the basic distribution form of the data. Therefore, it can be used for subsequent statistical feature extraction and feature fusion.

[0072] Second, in the feature extraction step, the skewness and kurtosis are calculated for different sensor data respectively, and the skewness and kurtosis are used as statistical features. The skewness calculation formula is as follows:

[0073]

[0074] n is the number of samples, x i is the i-th sample data, is the sample data mean, σ is the sample standard deviation;

[0075] The calculation formula for kurtosis is as follows:

[0076]

[0077] n is the number of samples, x i is the i-th sample data, is the sample data mean, and σ is the sample standard deviation.

[0078] At the actual code level, the skew and kurtosis functions in the scipy.stats library are used to calculate the skewness and kurtosis of normalized data.

[0079] Third, in the feature fusion step, the kurtosis data of multiple groups of sensors are collected and stored as a kurtosis feature data set. Similarly, the skewness data of multiple groups of sensors are stored as a skewness feature data set. A two-dimensional vector is formed by the kurtosis feature data and the kurtosis feature data set, in which one dimension represents skewness and the other dimension represents kurtosis.

[0080] Fourth, in the feature recognition step, the above two-dimensional vector is input through a pre-designed and trained CNN neural network model, and the result vector output by the neural network model can indicate the probability of abnormal skewness and / or kurtosis. Specifically, abnormal skewness means sensor failure or excessive temperature fluctuation, while abnormal kurtosis means that the abnormal value is too large. In this proposal, the basic structure of the CNN neural network algorithm model is not much different from that of other field applications, and both have convolutional layers, pooling layers, fully connected layers and output layers. Moreover, the training process is also a conventional training mode, that is, a sufficient amount of normal values ​​and abnormal values ​​are used for training. The only difference is that in the present invention, the sampling values ​​of the sensor are not used for training, but the skewness and kurtosis are calculated after normalization for training. In other words, the training process of the model is actually to train the model with two-dimensional vectors of normal skewness and kurtosis and two-dimensional vectors of abnormal skewness and kurtosis.

[0081] Among them, statistical analysis is used to perform statistical analysis on sensor data, including calculating statistical indicators such as mean, standard deviation, maximum value, minimum value, etc., to understand the distribution and change of data. Visual analysis is used to visualize sensor data in the form of charts, curves, etc., to intuitively show the changes and trends of data and help understand and analyze data.

[0082] In order to further improve the automation and intelligence level of environmental monitoring, the above-mentioned environmental detection patrol robot also includes an acoustic radar, which is communicatively connected to the above-mentioned controller. The above-mentioned acoustic radar is used to detect sound signals in the above-mentioned explosive warehouse. After obtaining the above-mentioned environmental information, the above-mentioned method also includes the following steps: obtaining the above-mentioned sound signal; performing voiceprint recognition on the above-mentioned sound signal to determine whether there is an explosion sound.

[0083] In this solution, the original sound signal may contain various sounds, including environmental noise, human voice, machine sound, and possible explosion sound. Voiceprint recognition is an audio signal processing technology used to identify and analyze the characteristics of sound signals. Voiceprint recognition is used to analyze the acquired sound signal to identify whether it contains a specific sound pattern, that is, an explosion sound. Through voiceprint recognition, it can be determined whether the acquired sound signal contains features that match the explosion sound. Through voiceprint recognition technology, it is possible to more accurately determine whether the sound signal contains an explosion sound, reducing the possibility of false positives and false negatives.

[0084] Specifically, the actual confirmation step may also include confirming the ambient sound information through an omnidirectional acoustic radar, and confirming whether there is an explosion sound and the direction of the sound source based on a voiceprint recognition algorithm.

[0085] In some embodiments, voiceprint recognition is performed on the sound signal to determine whether there is an explosion sound, which can be specifically achieved through the following steps: performing Fourier transform on the sound signal to obtain a spectrum diagram of the sound signal; taking the logarithm of the energy signal of the spectrum diagram to obtain a sound feature diagram; constructing a sound detection model, wherein the sound detection model is trained by a convolution algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes a historical sound feature diagram obtained within a historical time period, and a historical second detection result corresponding to the historical sound feature diagram, wherein the historical second detection result is used to characterize whether there is a historical explosion sound in the explosive warehouse; inputting the sound feature diagram into the sound detection model to obtain a second detection result corresponding to the sound feature diagram.

[0086] In this scheme, Fourier transform is a mathematical tool used to convert signals in the time domain (or space domain) into signals in the frequency domain, and to convert sound signals from the time domain to the frequency domain to obtain a sound spectrum diagram, which can extract key features from the sound signal. The spectrum diagram shows the energy distribution of the sound signal at different frequencies, which helps to identify the characteristics of the sound. Since the energy signal in the spectrum diagram may be very large or very small, direct processing may lead to numerical problems. Therefore, taking the logarithm of the energy signal is a common method, which can compress the numerical range and make the data more stable and easy to process. The sound feature diagram obtained in this way is more suitable for subsequent analysis and model training. By preprocessing the sound signal (such as Fourier transform and logarithmic transform), the influence of noise and redundant information can be reduced. The sound detection model is a machine learning model for identifying sound features, which can be used to detect explosion sounds in sound feature diagrams in real time without human intervention, thereby improving the level of automation and intelligence.

[0087] Specifically, the main steps of voiceprint recognition are as follows:

[0088] 1) Capture audio signals through omnidirectional acoustic radar;

[0089] 2) Calculate the spectrum of the audio signal through short-time Fourier transform, and take the logarithm of the energy signal of the spectrum to obtain the sound feature map;

[0090] 3) Input the sound feature map into the trained convolutional neural network model for recognition. The structure of the neural network model is an input layer, a convolution layer, a sampling layer, a fully connected layer and an output layer. The number of convolution kernels in the convolution layer is 10, and the sampling method of the sampling layer uses maximum sampling. The network uses the standard back propagation algorithm for learning, uses cross entropy as the cost function, and the output layer uses a softmax classifier.

[0091] The size and dimension of the input and output layers of the neural network model are fixed, but because the sound signal is highly random, the time length of each abnormal sound sample is not uniform, so the number of frames of each sample is different. If the time axis is compressed or expanded to ensure that the time length of each sample is the same, it is difficult to align them effectively and accurately. Therefore, this proposal uses a dynamic time warping method to warp all abnormal sound samples into the same number of frames to ensure that the feature parameters of each abnormal sound sample can be arranged into feature maps of the same size and input into the convolutional neural network.

[0092] Dynamic time warping is to build a correspondence between two sequence elements of different lengths according to the principle of shortest distance and evaluate the similarity of the two sequences. When building a correspondence between two sequence elements, the sequence needs to be extended or compressed. The calculation process is as follows:

[0093] 1) Define the distance measurement method between two time series. This proposal uses Euclidean distance, which can be replaced by Manhattan distance, cosine similarity, etc.

[0094] 2) Construct a cumulative distance matrix: Create a distance matrix by calculating the distance between any two points in the two time series;

[0095] 3) Calculate the cumulative distance matrix: Starting from the upper left corner of the distance matrix, gradually calculate the minimum distance to each point along the shortest path to form a cumulative distance matrix;

[0096] 4) Select the best path: Starting from the lower right corner of the cumulative distance matrix, trace back to find an optimal path, that is, the path with the minimum distance. The path with the minimum distance is the best time alignment path.

[0097] Use the best time alignment path to adjust the sound samples to the same number of frames. Simply put, the features of a sound sample are mapped to the frames of the standard sample through the best time alignment path, thereby achieving a uniform number of frames.

[0098] When training the neural network algorithm model for voiceprint recognition, samples are randomly selected from 70% of abnormal sound samples in five sub-libraries with different signal-to-noise ratios, and each training sub-set contains 6,300 samples. The test set contains five sub-test sets with different signal-to-noise ratios (no noise, 0dB, 5dB, 10dB, and 15dB), and 50% of abnormal sound samples are randomly selected from the remaining part of the five sub-libraries with different signal-to-noise ratios, and each sub-test set contains 1,350 samples. The validation set also contains five sub-validation sets with different signal-to-noise ratios (no noise, 0dB, 5dB, 10dB, and 15dB), and each sub-validation set contains 1,350 samples.

[0099] Furthermore, the inspection robot is controlled to go to the location where the abnormal sound is present, and the environment is confirmed using a temperature sensor, a particle sensor or a gas sensor, while image information is collected through a panoramic camera.

[0100] In some embodiments, after performing voiceprint recognition on the sound signal to determine whether there is an explosion sound, the method further includes the following steps: when it is determined that there is an explosion sound, determining the direction of the sound source based on sound detection technology, wherein the sound detection technology includes one or more of a microphone array method, an ultrasonic positioning method, and a spectrum analysis method.

[0101] In this solution, once the existence of the explosion sound is confirmed, sound detection technology will be used to determine the direction or position from which the sound is coming. Voiceprint analysis will be used to confirm the direction of the possible situation, reducing the possibility of false alarms and improving positioning accuracy.

[0102] Microphone array method: Use multiple microphones to form an array and determine the direction of the sound source by measuring the time difference or intensity difference when the sound reaches different microphones.

[0103] Ultrasonic positioning method: emit ultrasonic waves and receive the reflected signals, and determine the position of the sound (or object) by calculating the propagation time and direction of the ultrasonic waves.

[0104] Spectral analysis: By analyzing the spectral characteristics of the sound (such as frequency, amplitude, etc.), certain characteristics of the sound source can be inferred. Although it may not be directly used for positioning itself, it can provide useful clues to assist other positioning methods.

[0105] Specifically, this solution confirms possible abnormal conditions through anomaly analysis. During the confirmation process, the possible direction is mainly confirmed by acoustic radar, and then after comprehensive confirmation through video and other sensors, it is reported to the remote control center and formed into a visual chart.

[0106] The embodiment of the present application also provides an identification device for an environmental detection inspection robot. It should be noted that the identification device for the environmental detection inspection robot of the embodiment of the present application can be used to execute the identification method for the environmental detection inspection robot provided by the embodiment of the present application. The device is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware for a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0107] The following is an introduction to the identification device of the environmental detection inspection robot provided in the embodiment of the present application.

[0108] Figure 7 1 is a structural block diagram of an identification device of an environment detection inspection robot according to an embodiment of the present application. Figure 7 As shown, the device comprises:

[0109] A first acquisition unit 100 is used to acquire the above-mentioned environmental information, wherein the above-mentioned environmental information includes one or more of gas concentration, temperature, humidity and particle concentration;

[0110] A fusion unit 200 is used to fuse the plurality of the above-mentioned environmental information to obtain fused environmental information, wherein the fusion method includes one or more of a weighted average method, a Kalman filter method, a multi-Bayesian estimation method, a DS evidence reasoning method, a production rule, and a fuzzy logic reasoning method;

[0111] A construction unit 300 is used to construct an environment detection model, wherein the environment detection model is obtained by training a CNN algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes historical fusion environment information obtained in a historical time period and a historical first detection result corresponding to the historical fusion environment information, wherein the historical first detection result is used to characterize whether the historical environment in the explosive warehouse is abnormal;

[0112] The processing unit 400 is used to input the above-mentioned fused environment information into the above-mentioned environment detection model to obtain a first detection result corresponding to the above-mentioned fused environment information.

[0113] Through this embodiment, the inspection robot collects information on multiple environments of the explosive warehouse while moving in the explosive warehouse, and determines whether the environment in the explosive warehouse is abnormal through the paradigm of multi-data fusion, thereby avoiding the problem of low accuracy of single temperature detection. Environmental detection is achieved by collecting environmental information through multiple sensors, ensuring a high accuracy of the recognition results.

[0114] During the specific implementation process, the first acquisition unit includes an acquisition module, a first processing module, a second processing module and a third processing module. The acquisition module is used to acquire initial environmental information; the first processing module is used to perform mean filtering on the above initial environmental information to obtain first environmental information; the second processing module is used to perform normalization processing on the above first environmental information to obtain second environmental information; the third processing module is used to perform feature extraction on the above second environmental information to obtain the above environmental information.

[0115] In this scheme, the initial environmental information can be preprocessed, and the preprocessing includes mean filtering, normalization and feature extraction. Through mean filtering and normalization, the noise and redundant information in the original data can be removed, and the quality and reliability of the data can be improved. Normalization is to scale the data to a specific range (usually 0 to 1 or -1 to 1) to facilitate subsequent data analysis and processing. Feature extraction is the process of extracting the most representative information or pattern from the data. Feature extraction can reduce the dimension and complexity of the data and reduce the computational complexity of subsequent analysis and processing.

[0116] In order to further improve the automation and intelligence level of environmental monitoring, the above-mentioned environmental detection inspection robot also includes an acoustic radar, which is communicatively connected to the above-mentioned controller. The above-mentioned acoustic radar is used to detect the sound signal in the above-mentioned explosive warehouse. The above-mentioned device also includes a second acquisition unit and an identification unit. The second acquisition unit is used to acquire the above-mentioned sound signal after acquiring the above-mentioned environmental information; the identification unit is used to perform voiceprint recognition on the above-mentioned sound signal to determine whether there is an explosion sound.

[0117] In this solution, the original sound signal may contain various sounds, including environmental noise, human voice, machine sound, and possible explosion sound. Voiceprint recognition is an audio signal processing technology used to identify and analyze the characteristics of sound signals. Voiceprint recognition is used to analyze the acquired sound signal to identify whether it contains a specific sound pattern, that is, an explosion sound. Through voiceprint recognition, it can be determined whether the acquired sound signal contains features that match the explosion sound. Through voiceprint recognition technology, it is possible to more accurately determine whether the sound signal contains an explosion sound, reducing the possibility of false positives and false negatives.

[0118] In some embodiments, the recognition unit includes a fourth processing module, a fifth processing module, a construction module and a sixth processing module, the fourth processing module is used to perform Fourier transform on the above-mentioned sound signal to obtain a spectrum diagram of the above-mentioned sound signal; the fifth processing module is used to take the logarithm of the energy signal of the above-mentioned spectrum diagram to obtain a sound feature diagram; the construction module is used to construct a sound detection model, wherein the above-mentioned sound detection model is trained by a convolution algorithm using multiple sets of training data, and each set of training data in the above-mentioned multiple sets of training data includes a historical sound feature diagram obtained within a historical time period and a historical second detection result corresponding to the above-mentioned historical sound feature diagram, wherein the above-mentioned historical second detection result is used to characterize whether there is a historical explosion sound in the above-mentioned explosive warehouse; the sixth processing module is used to input the above-mentioned sound feature diagram into the above-mentioned sound detection model to obtain a second detection result corresponding to the above-mentioned sound feature diagram.

[0119] In this scheme, Fourier transform is a mathematical tool used to convert signals in the time domain (or space domain) into signals in the frequency domain, and to convert sound signals from the time domain to the frequency domain to obtain a sound spectrum diagram, which can extract key features from the sound signal. The spectrum diagram shows the energy distribution of the sound signal at different frequencies, which helps to identify the characteristics of the sound. Since the energy signal in the spectrum diagram may be very large or very small, direct processing may lead to numerical problems. Therefore, taking the logarithm of the energy signal is a common method, which can compress the numerical range and make the data more stable and easy to process. The sound feature diagram obtained in this way is more suitable for subsequent analysis and model training. By preprocessing the sound signal (such as Fourier transform and logarithmic transform), the influence of noise and redundant information can be reduced. The sound detection model is a machine learning model for identifying sound features, which can be used to detect explosion sounds in sound feature diagrams in real time without human intervention, thereby improving the level of automation and intelligence.

[0120] In some embodiments, the above-mentioned device also includes a determination unit, which is used to perform voiceprint recognition on the above-mentioned sound signal to determine whether there is an explosion sound. When it is determined that the above-mentioned explosion sound exists, the determination unit determines the direction of the sound source according to sound detection technology, wherein the above-mentioned sound detection technology includes one or more of a microphone array method, an ultrasonic positioning method and a spectrum analysis method.

[0121] In this solution, once the existence of the explosion sound is confirmed, sound detection technology will be used to determine the direction or position from which the sound is coming. Voiceprint analysis will be used to confirm the direction of the possible situation, reducing the possibility of false alarms and improving positioning accuracy.

[0122] The recognition device of the above-mentioned environment detection inspection robot includes a processor and a memory. The above-mentioned first acquisition unit, fusion unit, construction unit and processing unit are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize the corresponding functions. The above-mentioned modules are all located in the same processor; or, the above-mentioned modules are located in different processors in any combination.

[0123] The processor includes a kernel, and the kernel retrieves the corresponding program unit from the memory. One or more kernels can be set, and the kernel parameters can be adjusted to solve the problem that most robots for patrolling explosive warehouses in the prior art are equipped with a single temperature sensor to achieve environmental detection through temperature monitoring, and the accuracy of such recognition results is low.

[0124] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0125] An embodiment of the present invention provides a computer-readable storage medium, which includes a stored program, wherein when the program is running, the device where the computer-readable storage medium is located is controlled to execute the identification method of the environmental detection inspection robot.

[0126] An embodiment of the present invention provides a processor, and the processor is used to run a program, wherein the identification method of the environment detection inspection robot is executed when the program is running.

[0127] The present application also provides an environmental detection system, comprising one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include an identification method for executing any one of the above-mentioned environmental detection inspection robots.

[0128] An embodiment of the present invention provides a device, which includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, at least the following steps of the identification method of the environment detection inspection robot are implemented.

[0129] The devices in this article can be servers, PCs, PADs, mobile phones, etc.

[0130] A computer program product includes a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the identification method of the environmental detection inspection robot in each embodiment of the present application are implemented.

[0131] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0132] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0133] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0134] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0135] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0136] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0137] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0138] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0139] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0140] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0141] 1) The environmental detection inspection robot of the present application is a new inspection robot designed for use in explosive warehouses. By installing multiple sensors on the inspection robot, the inspection robot collects information on multiple environments of the explosive warehouse while moving in the explosive warehouse, thereby avoiding the problem of low accuracy of single temperature detection. Environmental detection is achieved by collecting environmental information through multiple sensors, ensuring a high accuracy of the recognition results.

[0142] 2) The identification method of the environmental detection inspection robot of the present application, while the inspection robot moves in the explosive warehouse, collects information on multiple environments of the explosive warehouse, and determines whether the environment in the explosive warehouse is abnormal through the paradigm of multi-data fusion, thereby avoiding the problem of low accuracy of single temperature detection, and realizes environmental detection by collecting environmental information through multiple sensors, ensuring a high accuracy of the identification result.

[0143] 3) The identification device of the environmental detection inspection robot of the present application collects information on multiple environments of the explosive warehouse while the inspection robot moves in the explosive warehouse, and determines whether the environment in the explosive warehouse is abnormal through the paradigm of multi-data fusion, thereby avoiding the problem of low accuracy of single temperature detection. Environmental detection is achieved by collecting environmental information through multiple sensors, ensuring a high accuracy rate of the identification result.

[0144] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. An environmental inspection robot, characterized in that: include: A sensing unit, comprising a plurality of sensors, wherein the sensing unit is used to detect environmental information in the explosive warehouse, wherein the environmental information includes at least one or more of gas concentration, temperature, humidity and particle concentration; A walking unit, used to enable the environmental detection inspection robot to move within the explosive warehouse; A controller is respectively connected to the sensing unit and the walking unit for communication. The controller is used to control the walking unit to work. The controller is also used to determine whether the environment in the explosive warehouse is abnormal according to the environmental information.

2. The environmental detection inspection robot according to claim 1, characterized in that: The sensing unit comprises: A gas sensor for detecting the gas concentration in the explosive warehouse; A temperature sensor, used to detect the temperature in the explosive warehouse; A humidity sensor, used for detecting the humidity in the explosive warehouse; The particle sensor is used to detect the particle concentration in the explosive warehouse.

3. The environmental detection inspection robot according to claim 1, characterized in that: The environmental detection inspection robot also includes: An acoustic radar is connected to the controller for communication, and is used for detecting sound signals in the explosive warehouse.

4. A method for identifying an environmental inspection robot according to any one of claims 1 to 3, characterized in that: The method further comprises: Acquiring the environmental information, wherein the environmental information includes one or more of gas concentration, temperature, humidity, and particulate matter concentration; Performing fusion processing on the plurality of environmental information to obtain fused environmental information, wherein the fusion method includes one or more of a weighted average method, a Kalman filter method, a multi-Bayesian estimation method, a DS evidence reasoning method, a production rule, and a fuzzy logic reasoning; Constructing an environment detection model, wherein the environment detection model is obtained by training with a CNN algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes historical fusion environment information obtained in a historical time period and a historical first detection result corresponding to the historical fusion environment information, wherein the historical first detection result is used to characterize whether the historical environment in the explosive warehouse is abnormal; The fused environment information is input into the environment detection model to obtain a first detection result corresponding to the fused environment information.

5. The method according to claim 4, characterized in that Obtaining the environment information includes: Get initial environment information; Performing mean filtering on the initial environment information to obtain first environment information; Normalizing the first environmental information to obtain second environmental information; Perform feature extraction on the second environmental information to obtain the environmental information.

6. The method according to claim 4, characterized in that The environment detection inspection robot further includes an acoustic radar, which is communicatively connected to the controller, and the acoustic radar is used to detect sound signals in the explosive warehouse. After acquiring the environment information, the method further includes: Acquiring the sound signal; Perform voiceprint recognition on the sound signal to determine whether there is an explosion sound.

7. The method according to claim 6, characterized in that Performing voiceprint recognition on the sound signal to determine whether there is an explosion sound, including: Performing Fourier transform on the sound signal to obtain a frequency spectrum of the sound signal; Taking the logarithm of the energy signal of the spectrum graph to obtain a sound feature graph; Constructing a sound detection model, wherein the sound detection model is trained by a convolution algorithm using multiple sets of training data, each set of training data in the multiple sets of training data includes a historical sound feature graph obtained in a historical time period and a historical second detection result corresponding to the historical sound feature graph, wherein the historical second detection result is used to characterize whether there is a historical explosion sound in the explosive warehouse; The sound feature graph is input into the sound detection model to obtain a second detection result corresponding to the sound feature graph.

8. The method according to claim 6, characterized in that After performing voiceprint recognition on the sound signal to determine whether there is an explosion sound, the method further includes: When it is determined that the explosion sound exists, the direction of the sound source is determined based on sound detection technology, wherein the sound detection technology includes one or more of a microphone array method, an ultrasonic positioning method and a spectrum analysis method.

9. An identification device based on the environment detection inspection robot according to any one of claims 1 to 3, characterized in that: The device also includes: A first acquisition unit, configured to acquire the environmental information, wherein the environmental information includes one or more of gas concentration, temperature, humidity and particle concentration; A fusion unit, used for fusing the plurality of environmental information to obtain fused environmental information, wherein the fusion method includes one or more of a weighted average method, a Kalman filter method, a multi-Bayesian estimation method, a DS evidence reasoning method, a production rule, and a fuzzy logic reasoning method; A construction unit is used to construct an environment detection model, wherein the environment detection model is obtained by training a plurality of sets of training data through a CNN algorithm, wherein each set of training data in the plurality of sets of training data includes historical fusion environment information obtained in a historical time period and a historical first detection result corresponding to the historical fusion environment information, wherein the historical first detection result is used to characterize whether the historical environment in the explosive warehouse is abnormal; The processing unit is used to input the fused environment information into the environment detection model to obtain a first detection result corresponding to the fused environment information.

10. An environmental detection system, characterized in that: include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a recognition method for executing the environmental detection inspection robot described in any one of claims 4 to 8.

Images