Fire extinguisher performance test method and system and computer program product

Abstract

The invention relates to the technical field of fire extinguishing testing, in particular to a fire extinguisher performance testing method and system and a computer program product.The method comprises the steps that according to the applicability of a target fire extinguisher, a target fire scene environment is simulated on the basis of environmental factors having influences on combustion and fire extinguishing effects; according to a preset control mode of the fire extinguishing robot, the fire extinguishing robot is controlled to execute fire extinguishing operation on the target fire scene environment through the target fire extinguisher; and in the fire extinguishing operation execution process, relevant information reflecting the fire extinguishing effect of the target fire extinguisher is collected in real time, so that a performance test result of the target fire extinguisher is generated. Therefore, the problems that in the prior art, due to the fact that the detection result is greatly influenced by factors such as experience, techniques and states of fire extinguishing experiment operators and the temperature, humidity and ventilation degree of a testing environment, performance testing errors are large possibly, the operators need to frequently approach a fire source, and safety risks exist are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fire extinguishing test, and particularly relates to a fire extinguisher performance test method, system and computer program product. BACKGROUND

[0002] In the related art, after the fire extinguishing sample and the fire extinguishing model are prepared, an operator wears a protective suit, holds a fire extinguisher and sprays according to the requirements of the standard method, so that the fire extinguishing operation is effectively performed, the fire extinguishing effect of the fire extinguisher can be clearly shown, subsequent observation and evaluation are facilitated, and scientific evaluation of the performance of the fire extinguisher is achieved.

[0003] However, in the related art, the detection result is greatly affected by the experience, method, state of the fire extinguishing experiment operator and factors such as the temperature, humidity and ventilation degree of the test environment, which may result in a large performance test error and reduce the reliability of the detection result. In addition, the operator needs to frequently approach the fire source, which poses a potential safety risk and needs to be improved. SUMMARY

[0004] The present application provides a fire extinguisher performance test method, system and computer program product to solve the problems in the related art that the detection result is greatly affected by the experience, method, state of the fire extinguishing experiment operator and factors such as the temperature, humidity and ventilation degree of the test environment, which may result in a large performance test error, the operator needs to frequently approach the fire source and poses a safety risk.

[0005] The first aspect embodiment of the present application provides a fire extinguisher performance test method, comprising the following steps: simulating a target fire field environment based on environmental factors that have an impact on combustion and fire extinguishing effect according to the applicability of a target fire extinguisher; controlling a fire extinguishing robot to perform fire extinguishing operation on the target fire field environment using the target fire extinguisher according to a preset control mode of the fire extinguishing robot; and collecting relevant information reflecting the fire extinguishing effect of the target fire extinguisher in real time during the fire extinguishing operation to generate a performance test result of the target fire extinguisher.

[0006] Optionally, in an embodiment of the present application, the control mode includes a manual control mode; and the method comprises: in response to the preset control mode of the fire extinguishing robot being the manual control mode, controlling the fire extinguishing robot to perform fire extinguishing operation on the target fire field environment using the target fire extinguisher according to a manual fire extinguishing operation instruction issued by an operator.

[0007] Optionally, in an embodiment of the present application, the control mode includes an automatic control mode; the method further comprises: in response to the preset control mode of the fire extinguishing robot being the automatic control mode, importing a standard fire extinguishing operation model matched with the target fire extinguisher; performing model analysis on the standard fire extinguishing operation model to obtain standard operation information of the fire extinguishing robot; and controlling the fire extinguishing robot to perform fire extinguishing operation on the target fire field environment using the target fire extinguisher according to the standard operation information of the fire extinguishing robot.

[0008] Optionally, in an embodiment of the present application, the control mode includes an automatic intelligent control mode; the method further comprises: in response to the preset control mode of the fire extinguishing robot being the automatic intelligent control mode, collecting feedback information after the fire extinguishing robot performs current fire extinguishing operation using the target fire extinguisher; adjusting behavior information of the fire extinguishing robot according to the feedback information, and performing fire extinguishing operation on the target fire field environment based on the adjusted behavior information of the fire extinguishing robot.

[0009] Optionally, in an embodiment of the present application, the feedback information includes environment state information, combustion state information, fire extinguishing robot state information, fire extinguishing robot limb state information, fire extinguisher state information, and cost generated in the fire extinguishing process; and the adjusting the behavior information of the fire extinguishing robot according to the feedback information comprises: serializing and encoding the environment state information, the combustion state information, the fire extinguishing robot state information, the fire extinguishing robot limb state information, the fire extinguisher state information, and the cost generated in the fire extinguishing process to obtain an input feature vector; inputting the input feature vector into a fire extinguishing robot behavior driving model to predict and obtain behavior information of the fire extinguishing robot at the next moment; importing expert demonstration behavior information from a pre-constructed expert demonstration data set, and comparing the expert demonstration behavior information with the predicted and obtained behavior information of the fire extinguishing robot at the next moment, to minimize a pre-constructed total loss function and minimize the injury degree of the fire extinguishing robot as a constraint, to perform parameter optimization on the fire extinguishing robot behavior driving model; using the parameter-optimized fire extinguishing robot behavior driving model to predict and obtain the behavior information of the fire extinguishing robot at the next moment, to obtain the adjusted behavior information of the fire extinguishing robot; wherein the fire extinguishing robot driving model is constructed based on a multi-head attention mechanism neural network model; and the multi-head attention mechanism neural network model includes an encoder and a decoder, the encoder is used to encode the input feature vector to obtain an encoded feature vector, and the decoder is used to perform decoding operation according to the encoded feature vector and the input feature vector to predict and obtain the behavior information of the fire extinguishing robot at the next moment.

[0010] Optionally, in an embodiment of the present application, the expression of the total loss function is as follows:

[0011] ; ; wherein, represents the total loss function; describes all internal fire-fighting robot activity parameters; N represents the number of expert demonstration data sets; M represents the number of actual running data sets of the fire-fighting robot; represents the expert demonstration data set, which contains a number of complete expert demonstration action trajectories; represents the expert demonstration behavior information; represents the predicted obtained behavior information of the fire-fighting robot at the next time; , represents the trajectory loss cost of ; represents the average loss cost of ; represents the kth step (time dimension) of the fire-fighting experiment (including the environment, robot and fire extinguisher state) state; represents the kth step of the fire-fighting robot action; represents the weighted loss cost caused by the state and the fire-fighting robot action ; represents the generation probability of , , is the initial state distribution ; is the initial state of the fire-fighting experiment; K represents the total number of trajectory paths in the entire fire-fighting process; represents the probability of selecting the fire-fighting robot action under the state ; represents the probability of transferring to the state after selecting the fire-fighting robot action under the state ; represents the k+1th step of the fire-fighting experiment state; represents the trajectory loss cost of ; represents the relative preference for other trajectories; represents the trajectory negative cost of ; represents the injury state of the kth step of the fire-fighting robot, , respectively represent the scalding, choking and collision of the fire-fighting robot; represents the safety loss weight.

[0012] Optionally, in one embodiment of the present application, the expressions of the serialization encoding of the environment state information, the combustion state information, the fire extinguishing robot state information, the fire extinguishing robot limb state information, the fire extinguisher state information and the cost generated by the fire extinguishing process are as follows:

[0013]

[0014] wherein, is the fire extinguishing experiment state vector; is the Embedding sequence encoding function; represents the splicing operation, which splices all the state information into a complete vector; and are the data vectorized representations of the environment state information, the combustion state information, the fire extinguishing robot state information, the fire extinguishing robot limb state information, the fire extinguisher state information and the cost generated by the fire extinguishing process, respectively; represents the encoding of the time dimension;

[0015] wherein, represent the environment temperature, the environment humidity and the environment smoke concentration, respectively;

[0016] represent the combustion temperature, the combustion material type and the combustion range, respectively;

[0017] represent the visual information, the auditory information, the tactile information, the olfactory information, the perceived temperature information, the motion information, the position information and the remaining consumable energy of the fire extinguishing robot, respectively;

[0018] represent the body position and orientation, the head position and orientation, the arm position and orientation and the finger position and orientation of the fire extinguishing robot, respectively;

[0019] ,​​​​​​​​​​​​​​​​​​​​​ respectively represent the type, weight, internal pressure, position, angle of the fire extinguisher, the position of the pull valve, the angle of the pull valve, the spray speed of the nozzle, and the remaining fire extinguishing agent mass.

[0020] Optionally, in an embodiment of the present application, the relevant information reflecting the fire extinguishing effect of the target fire extinguisher at least includes the burning state information and the injury degree of the fire extinguishing robot; and the generating of the performance test result of the target fire extinguisher comprises: in response to the injury degree of the fire extinguishing robot exceeding a set threshold, generating a test performance report of a fire extinguishing test failure; and in response to the injury degree of the fire extinguishing robot not exceeding the set threshold and the burning state information satisfying a fire extinguishing success judgment standard, generating a test performance report of a fire extinguishing test success.

[0021] The second aspect embodiment of the present application provides a fire extinguisher performance test system, comprising: a fire scene simulation module, configured to simulate a target fire scene environment based on environmental factors that have an influence on burning and fire extinguishing effect according to the applicability of a target fire extinguisher; a robot control module, configured to control a fire extinguishing robot to perform a fire extinguishing operation on the target fire scene environment using the target fire extinguisher according to a preset control mode of the fire extinguishing robot; and a result determination module, configured to collect relevant information reflecting the fire extinguishing effect of the target fire extinguisher in real time during the execution of the fire extinguishing operation, so as to generate a performance test result of the target fire extinguisher.

[0022] The third aspect embodiment of the present application provides a computer program product, comprising computer programs / instructions, which are executed by a processor to implement the fire extinguisher performance test method as above.

[0023] The embodiments of the present application can simulate a target fire scene environment based on environmental factors that have an influence on burning and fire extinguishing effect according to the applicability of a target fire extinguisher, and then control a fire extinguishing robot to perform a fire extinguishing operation on the target fire scene environment using the target fire extinguisher according to a preset control mode of the fire extinguishing robot, so as to collect relevant information reflecting the fire extinguishing effect of the target fire extinguisher in real time during the execution of the fire extinguishing operation, thereby realizing the automatic output of a standardized fire extinguisher performance test result, effectively automating the mechanical repetitive labor in the fire extinguishing experiment, and helping to improve the accuracy and safety of the fire extinguisher fire extinguishing performance test. Thus, the problems in the related art that the detection result is greatly influenced by the experience, method, state of the fire extinguishing experiment operator, and factors such as the temperature, humidity, and ventilation degree of the test environment, which may lead to a large performance test error, and the operator needs to frequently approach the fire source, which poses a safety risk, and the like, are solved. ​​​​​​

[0024] Additional aspects and advantages of the application will be set forth in part in the description which follows, and in part will become apparent to those skilled in the art upon examination of the following and / or can be learned by practice of the application. BRIEF DESCRIPTION OF DRAWINGS

[0025] The above and / or additional aspects and advantages of the application will become apparent and be more readily understood through reference to the following description, taken in conjunction with the accompanying drawings, wherein:

[0026] Figure 1 A flow chart of a method for testing performance of a fire extinguisher according to an embodiment of the application;

[0027] Figure 2 A schematic diagram of a framework structure of a method for testing performance of a fire extinguisher according to an embodiment of the application;

[0028] Figure 3 A flow chart of a method for manually controlling a fire extinguisher to extinguish a fire according to an embodiment of the application;

[0029] Figure 4 A flow chart of a method for automatically controlling a fire extinguisher to extinguish a fire according to an embodiment of the application;

[0030] Figure 5 A flow chart of a method for updating a fire extinguishing model according to an embodiment of the application;

[0031] Figure 6 A schematic diagram of a structure of a system for testing performance of a fire extinguisher according to an embodiment of the application. DETAILED DESCRIPTION

[0032] Embodiments of the application are described in detail below with reference to the attached drawing figures, wherein the same or like reference numerals are used throughout the figures to refer to same or like elements or elements with same or similar function. The embodiments described below are exemplary and are intended to be illustrative of the application and are not to be construed as limiting the application.

[0033] A performance testing method of a fire extinguisher according to an embodiment of the present application is described below with reference to the accompanying drawings. In view of the problems in the related art mentioned above, that is, the detection result is greatly affected by the experience, method, and state of the fire extinguishing experiment operator, and the temperature, humidity, and ventilation degree of the test environment, which may result in a large performance testing error, and the operator needs to frequently approach the fire source, which is a safety risk, the present application provides a performance testing method of a fire extinguisher. In the method, the target fire field environment can be simulated based on environmental factors that have an impact on combustion and fire extinguishing effect according to the applicability of the target fire extinguisher. Then, the target fire extinguisher is used to perform fire extinguishing operation on the target fire field environment according to the preset control mode of the fire extinguishing robot. During the performance of the fire extinguishing operation, relevant information reflecting the fire extinguishing effect of the target fire extinguisher can be collected in real time, so as to automatically output a standardized fire extinguisher performance testing result. Thus, the mechanical repetitive labor in the fire extinguishing experiment is effectively automated, which helps to improve the accuracy and safety of the fire extinguisher performance testing.

[0034] Specifically, Figure 1 A flowchart of a performance testing method of a fire extinguisher according to an embodiment of the present application is shown in FIG. 1.

[0035] As Figure 1 shown, the performance testing method of the fire extinguisher includes the following steps:

[0036] In step S101, the target fire field environment is simulated based on environmental factors that have an impact on combustion and fire extinguishing effect according to the applicability of the target fire extinguisher.

[0037] It can be understood that the target fire extinguisher refers to the fire extinguisher that needs to be tested for performance, the environmental factors include but are not limited to the combustion type, humidity environment, temperature environment, ventilation condition, and combustion state, and the target fire field environment refers to different combustion environments arranged according to the applicability of the fire extinguisher that needs to be tested for performance.

[0038] In actual execution, as Figure 2 shown, the fire field scene can be simulated to set the flame effect caused by various materials. The combustion materials are prepared before the fire extinguishing experiment starts, and different combustion environments can be arranged according to the applicability of different fire extinguishers. In this way:

[0039] The combustion type includes but is not limited to electrical fire, gas fire, metal fire, and chemical fire.

[0040] The humidity environment includes but is not limited to dry, humid, and normal.

[0041] Temperature environment includes but is not limited to normal temperature 25℃, high temperature 40℃, low temperature 0℃;

[0042] Ventilation conditions include but are not limited to closed space, wind level low / medium / high;

[0043] Combustion state includes but is not limited to duration, temperature, flammable object quantity, and explosive degree;

[0044] Further, the embodiments of the present application can be provided with various types of sensors for real-time monitoring of fire information data and synchronously giving the intelligent data control module.

[0045] The embodiments of the present application can simulate the target fire environment based on environmental factors that have an impact on combustion and fire extinguishing effect according to the applicability of the target fire extinguisher, which is conducive to ensuring the standardization of test conditions, avoiding the uncontrollable interference of real fire with experimental results, and supporting repeated experiments, which helps to quickly verify the performance of the fire extinguisher and improve test efficiency.

[0046] In step S102, according to the preset control mode of the fire extinguishing robot, the fire extinguishing robot performs fire extinguishing operation on the target fire environment using the target fire extinguisher.

[0047] It can be understood that the preset control mode includes manual control mode, automatic control mode and automatic intelligent control mode.

[0048] Specifically, as shown in Figure 2 The embodiments of the present application simulate the fire extinguishing operation of real fire extinguishing experiment personnel in the fire range by the fire extinguishing robot, including:

[0049] Setting a simulation mechanical arm to simulate the operator's arm for simulating the action of holding the fire extinguisher, such as moving the position, angle and nozzle direction of the fire extinguisher;

[0050] Setting a simulation finger to simulate the operator's finger action for pulling the action of controlling the fire extinguisher valve, the mechanical arm can accurately control the fire extinguisher within the freedom of the human arm and pull the control fire valve within the range of human finger strength to complete the fire extinguishing operation;

[0051] Setting a mechanical leg or universal wheel for simulating the leg movement of a person according to visual feedback information, freely moving within the effective fire extinguishing range around the fire;

[0052] Setting a camera to simulate the human eye for real-time capturing of position state and flame state, etc.

[0053] The robot simulation flaw detection sensor is arranged to collect the scalding, choking and collision conditions of the fire extinguishing robot according to the sensors arranged at the key positions of the robot, including but not limited to the hand, eye, mouth and nose. The scalding condition can be quantitatively characterized by collecting temperature information through the temperature sensor, the choking condition can be quantitatively characterized by the smoke concentration collected through the smoke concentration sensor, and the collision condition can be quantitatively characterized by collecting pressure information through the pressure sensor. It should be understood that the smaller the collected temperature, smoke concentration and collision pressure, the smaller the injury degree of the corresponding fire extinguishing robot. The collected temperature, smoke concentration and collision pressure can jointly represent the injury degree of the fire extinguishing robot.

[0054] Further, as shown in Figure 2 The robot control module can be used to control the robot through a remote communication mode, including manual control, automatic control and automatic intelligent control.

[0055] Optionally, in one embodiment of the present application, it comprises: in response to the preset control mode of the fire extinguishing robot being a manual control mode, controlling the fire extinguishing robot to perform fire extinguishing operation on the target fire field environment by using the target fire extinguisher according to the manual fire extinguishing operation instruction issued by the operator.

[0056] Specifically, as shown in Figure 3 The manual control operation method replaces the direct artificial fire extinguishing operation, and initializes the instruction set data of the fire extinguishing model, including the following steps:

[0057] Step 1: Prepare various important equipment required for the experiment in the initial stage:

[0058] The fire extinguishers participating in the experiment are of various types (dry powder fire extinguishers, foam fire extinguishers, water-based fire extinguishers, carbon dioxide fire extinguishers and clean gas fire extinguishers), and each type of fire extinguisher can act on different types of fires. Therefore, different fire extinguishing combustion models can be constructed for different types of fires. For example, an A-class fire combustion model is needed to test the dry powder fire extinguisher, and a B-class fire combustion model is needed to test the foam fire extinguisher.

[0059] The basic environment for simulating combustion is prepared, for example: ventilation area 4m2, relative humidity 40%, environmental temperature 25℃, oxygen concentration 40%;

[0060] Prepare the fire extinguishing robot, which is equipped with a mechanical arm and a mechanical finger, can hold and use the fire extinguisher through remote control (including manual control), and can move freely in the fire extinguishing site;

[0061] Prepare temperature, humidity, smoke concentration and other sensors to monitor the state of the combustion environment, the combustion model and the key positions of the fire extinguishing robot;

[0062] Prepare other key elements, including the robot controller cut into manual mode, real-time monitoring module open, and real-time synchronization data to the results module and intelligent data control module.

[0063] Step 2, ignite the fire extinguishing combustion model, and control the fire extinguishing robot by manual control mode to extinguish the fire, at the same time, open the fire extinguishing robot itself's damage sensor, and synchronize the fire extinguishing robot control information and damage information to the robot control module for simulation feedback and real-time monitoring.

[0064] Step 3, after the fire extinguishing robot is manually controlled to extinguish the fire successfully, manually end the fire extinguishing experiment.

[0065] Step 4, the real-time monitoring module synchronizes the result data to the result determination module and the intelligent data control module.

[0066] Step 5, the result determination module summarizes the experimental report according to the combustion model, experimental environment, robot damage degree and other data, and automatically outputs the report as a standard document for the inspector to review.

[0067] Step 6, the intelligent control model module collects the manual operation process in the actual state, initializes the instruction set data of the fire extinguishing model, and is used to initialize the automatic control fire extinguishing robot extinguishing process.

[0068] Optionally, in an embodiment of the present application, it comprises: in response to the preset control mode of the fire extinguishing robot being automatic control mode, importing a standard fire extinguishing operation model matched with the target fire extinguisher; performing model analysis on the standard fire extinguishing operation model to obtain standard operation information of the fire extinguishing robot; and controlling the fire extinguishing robot to perform fire extinguishing operation on the target fire field environment by using the target fire extinguisher according to the standard operation information of the fire extinguishing robot.

[0069] It can be understood that, by collecting multiple complete manual control results and solidifying them into the fire extinguishing model as preset fixed operation trajectories, the embodiment of the present application realizes fire extinguishing operation by automatically performing the predetermined fire extinguishing operation.

[0070] Specifically, as shown in Figure 4 , the embodiment of the present application builds an unsupervised feedback model in the intelligent automatic fire extinguishing experiment by considering the influence factors such as robot damage and environmental pollution during the fire extinguishing process, and standardizes the fire extinguishing process automatically, including the following steps:

[0071] Step 1, prepare various important equipment needed for the experiment in the initial stage:

[0072] The fire extinguishers participating in the experiment are of various types (dry powder extinguishers, foam extinguishers, water-based extinguishers, carbon dioxide extinguishers, and clean gas extinguishers), and each type of extinguisher can act on different types of fires;

[0073] The fire extinguishing combustion model corresponding to the fire extinguisher (including wood, oil, metal, chemical, etc.), for example, testing dry powder extinguishers requires building a class A fire combustion model, and testing foam extinguishers requires selecting a class B fire combustion model;

[0074] The basic environment for simulating combustion, for example: ventilation area 4㎡, relative humidity 40%, ambient temperature 25℃, oxygen concentration 40%;

[0075] Prepare the fire extinguishing robot, the required fire extinguishing robot is equipped with a mechanical arm and mechanical fingers, which can hold and use the fire extinguisher through remote control (including manual control mode) and can move freely in the fire scene;

[0076] Prepare temperature, humidity, smoke concentration, etc. Sensors are used to monitor the state of the combustion environment, combustion model, and key parts of the fire extinguishing robot;

[0077] Prepare other key elements, including importing the robot controller into the automatic control mode, starting the real-time monitoring module, and synchronizing real-time data to the result determination module and intelligent data control module.

[0078] Step 2, ignite the fire extinguishing combustion model, and control the fire extinguishing robot to extinguish the fire through the automatic control mode, while starting the robot's own damage detection sensor, and synchronizing the robot control information and damage information to the robot control module for simulation feedback and real-time monitoring.

[0079] Step 3, the real-time monitoring module transmits the data in the process to the result determination module in real time.

[0080] Step 4, the result determination module determines whether the experiment can continue based on the combustion state and the robot state. If the robot's damage state is determined to be serious, go to step 5, if the temperature of the combustible material is reduced below the critical point, go to step 6, otherwise, return to step 3 to continue the experiment.

[0081] Step 5, the result determination module determines that the fire extinguishing experiment has failed, exits the experiment, and outputs a fire extinguishing experiment failure report.

[0082] Step 6, the result determination module determines that the fire extinguishing experiment is successful, summarizes the experiment report based on the combustion model, experimental environment, and robot damage degree, etc. data, and automatically outputs it as a standard document for the inspector to review.

[0083] Step 7, the real-time monitoring module synchronizes the result state to the intelligent data control module.

[0084] Step 8, the intelligent data control module adjusts the automatic instruction set of the fire extinguishing model according to the result state, including environmental information, robot injury state and other information as negative feedback parameters, for the next automatic experiment operation.

[0085] Optionally, in an embodiment of the present application, it comprises: in response to the preset control mode of the fire extinguishing robot being an automatic intelligent control mode, collecting feedback information after the fire extinguishing robot performs the current fire extinguishing operation using the target fire extinguisher; adjusting the behavior information of the fire extinguishing robot according to the feedback information, and performing the fire extinguishing operation on the target fire scene environment based on the adjusted behavior information of the fire extinguishing robot.

[0086] It can be understood that the embodiments of the present application can collect feedback information (including robot injury information and fire extinguishing effect information) collected in the process of automatically controlling fire extinguishing by the fire extinguishing model, and perform self-learning adjustment on the fire extinguishing operation (moving route, fire extinguisher position and angle, fire extinguishing valve pulling force, fire extinguishing nozzle direction angle, and fire extinguisher spraying time), complete the adjustment of the fire extinguishing model through a large number of experiments, realize the intelligent fire extinguishing function, and then can perform the fire extinguishing operation through the automatic intelligent control mode.

[0087] Specifically, in the data initialization phase of the fire extinguishing model, the experimental expert and the fire extinguisher configuration and installation data acquisition sensor are collected, and the captured experimental expert position and action information, fire extinguisher state information and the like are input into the attention neural network as input data in the actual experiment process, wherein, in combination with the fire extinguishing model, the specific steps are as follows: Figure 5 As shown in the figure, the specific steps are as follows:

[0088] Step 1: data acquisition.

[0089] 1. The data collected by the self-learning AI neural network includes three parts:

[0090] (1) The following information is captured through intelligent wearable devices (including but not limited to: smart glasses, smart earphones, smart breathing devices, position sensors, motion sensors, skin temperature sensors, and skin pressure sensors):

[0091] 1) Visual information: the collected video data is imaged, and the color, brightness, depth and other dimensional information of the image are input as visual feature vectors;

[0092] 2) Auditory information: the collected sound information is split into left and right channels, and the sound wave signal or audio feature is digitized as an auditory feature vector;

[0093] 3) Tactile information: The array collected from the pressure sensor is input as a multi-dimensional vector (each set point as a dimension) as a tactile feature vector input;

[0094] 4) Olfactory information: The concentration data collected from the chemical sensor or gas sensor is represented as an olfactory feature vector input;

[0095] 5) Perceived temperature information: The information collected from the temperature sensor is quantified as a multi-dimensional vector as a temperature perception feature vector input;

[0096] 6) Motion information: The information collected from the motion information sensor is split into individual dimension data, including but not limited to head, leg, arm, waist, finger, as a motion feature vector input;

[0097] 7) Position information: By collecting information from the accelerometer, gyroscope and magnetometer, the motion direction and attitude angle of the robot are calculated; by collecting the timestamp information from the UWB anchor point, the position information of the robot in the fire extinguishing experiment field is calculated; and the position and motion information are input as state feature vectors;

[0098] 8) Remaining consumable energy: The remaining consumable energy of the robot in the fire extinguishing experiment field is collected as an energy feature vector input.

[0099] (2) Grasp by motion state sensor installed on the fire extinguisher;

[0100] 1) Ontology information: The motion and attitude information of the fire extinguisher is digitized, and the position, angle, motion speed, etc. of the fire extinguisher ontology are input as state feature vectors;

[0101] 2) Nozzle state: The angle, valve force angle and size, etc. of the fire extinguisher nozzle information vector is input as a feature vector;

[0102] 3) Spray speed information: The speed and volume information of the material (including but not limited to gas, fluid, particle cluster) sprayed out of the fire extinguisher nozzle is input as a feature vector;

[0103] (3) Collect environmental information and combustion information by setting various types of sensors;

[0104] 1) Environmental information: Environmental temperature, environmental humidity and environmental smoke concentration, etc. as feature vector input.

[0105] 2) Combustion information: Combustion temperature, combustion material type and combustion range, etc. as feature vector input.

[0106] 2. Data labeling, including labeling the fire extinguishing process of fire extinguishing experiment experts after data collection is completed, and standardizing the disassembly of actions:

[0107] (1) Stage labeling: preparation stage, holding fire extinguisher, observing burning state, approaching fire source, avoiding flame, avoiding smoke, extinguishing preparation, posture adjustment, spraying fire extinguishing, extinguishing success determination, and ending fire extinguishing;

[0108] (2) Action labeling: moving towards the fire source, moving away from the fire source, moving away from the smoke, holding the fire extinguisher, aiming the fire extinguishing nozzle at the fire source, pulling the fire extinguisher valve, point spraying fire extinguishing, sweeping fire extinguishing, and extinguishing fire around the fire source;

[0109] (3) Feedback labeling: risk of burns, risk of choking smoke, and risk of collision.

[0110] Step 2: Imitation learning.

[0111] The obtained data is used as expert demonstration behavior, the data is divided into state and action dimensions, AI intelligent model is used for behavior cloning, and AI training model is initialized.

[0112] 1. Quantitative model: vectorize the collected data and organize it into demonstration model data, including:

[0113] Environmental state: ;

[0114] Burning state: ; Among them, the type of burning material In actual calculation, it can be but not limited to quantified by its corresponding physical and chemical parameters (such as emission factor, carbon content, heat value);

[0115] Robot state:

[0116] ;

[0117] Robot limb state: ;

[0118] Fire extinguisher state information: ;

[0119] Robot limb action: ; Among them, , , , , , , , , , , , , respectively represent the movement direction and movement acceleration of the body, head, arm, finger and fire extinguisher, and the size and angle of the force on the fire extinguisher pull valve;

[0120] Robot injury state: ; , , Respectively represent the scalding, choking and collision of the fire-fighting robot;

[0121] Among them, the state , , , , Respectively represent the current environment, burning, robot (including limbs) and fire extinguisher state, which are input information; the predicted robot limb action is output, used to perform specific actions; As feedback, used to correct the specific action of the robot fire extinguishing.

[0122] 2, Initialize the simulation learning model (including but not limited to IRL-Inverse Reinforcement Learning), internalize the collected demonstration actions into the AI model, and the loss function is described as:

[0123] ; Among them, Indicates the total loss function; Describes all internal fire-fighting robot activity parameters; N represents the number of expert demonstration data sets; M represents the number of fire-fighting robot actual running data sets; Indicates the expert demonstration data set, which contains several complete expert demonstration action trajectories; Indicates the expert demonstration behavior information; Indicates the predicted behavior information of the fire-fighting robot at the next time; , indicates the trajectory loss cost of ; Indicates the average loss cost of ; Indicates the k-th step of the fire-fighting experiment state; Indicates the k-th step of the fire-fighting robot action; Indicates the weighted loss cost caused by the state and the fire-fighting robot action ; Indicates the generation probability of , , is the initial state distribution ; is the initial state of the fire extinguishing experiment; K represents the total number of trajectory paths of the entire fire extinguishing process; represents the state selects the probability of the action of the fire extinguishing robot ; represents the state selects the probability of the action of the fire extinguishing robot after which it is transferred to the state ; represents the state of the fire extinguishing experiment at the k+1 step; represents the trajectory loss cost, including the consumption of time and energy; represents the relative preference for other trajectories; represents the negative cost of the trajectory; represents the injury state of the fire extinguishing robot at the k step, , respectively represent the scalding, choking, and collision of the fire extinguishing robot; represents the safety loss weight.

[0124] 3. Constructing a neural network model architecture (including but not limited to Transformer)

[0125] Embedding serialization encoding input:

[0126] ; wherein, is the fire extinguishing experiment state vector; is an Embedding sequence encoding function; represents a splicing operation that splices all state information into a complete vector; , , , , and are the environmental state information, the burning state information, the fire extinguishing robot state information, the fire extinguishing robot limb state information, the fire extinguisher state information, and the cost generated by the fire extinguishing process, respectively; including the time and energy cost consumed by the fire extinguishing process;

[0127] Encoder encoding:

[0128] ; wherein, is a Multi-Head Attention-based feedforward network;

[0129] Decoder decoding:

[0130] wherein, is the serialized input of the previous round of historical actions;

[0131] Loss-driven parameter optimization;

[0132] 1. Forward propagation: input information is generated through the encoder-decoder to generate a predicted action

[0133] 2. Loss calculation: using to compare the predicted action with the expert trajectory , combined with feedback to correct the parameters;

[0134] 3. Backpropagation and parameter update:

[0135] Step 3: drive the robot.

[0136] Steps 4-6: update the reward results through the loss function, and then adjust the robot behavior and adjust the fire extinguishing model.

[0137] Step 7: output the final fire extinguishing model result.

[0138] The embodiment of the present application initializes the model data by collecting real fire extinguishing scene information, and adjusts the data model with manual control assistance to generate an initial corresponding state corresponding to the complete fire extinguishing automatic operation steps of the fire extinguisher; as the initialization data of the fire extinguishing model, the flame state, the fire extinguishing environment, the robot injury state and other data are used as result feedback data to form a self-learning feedback training model. Through repeated training, the fire extinguishing model is adjusted to the best state, aiming to output a series of operation information such as position, angle, pulling force, spraying direction, etc. required when the corresponding fire extinguisher reaches the best fire extinguishing effect, and collect the effective fire extinguishing range as the basis for comprehensive judgment of the fire extinguishing effect, guiding the fire extinguishing behavior.

[0139] The embodiment of the present application reduces the influence of human factors on the test results through automation and intelligent technology, reduces the risk of fire experiment personnel contacting the fire source, improves the operation safety, and uses artificial intelligence technology to learn the negative feedback results of the fire extinguishing experiment, unsupervised self-learning fire extinguishing operation, and effectively improves the test efficiency.

[0140] In step S103, during the execution of the fire extinguishing operation, relevant information reflecting the fire extinguishing effect of the target fire extinguisher is collected in real time to generate the performance test result of the target fire extinguisher.

[0141] ​​​Specifically, as shown in Figure 2 According to the fire extinguishing result output module, after the fire extinguishing action is completed, a comprehensive determination is made on a series of result data such as the flame temperature and the smoke concentration, the fire extinguishing effect data is determined, and a complete fire extinguishing test report is automatically output based on the data. The report that does not meet the standard is adjusted and output by the intelligent report output model after the feedback of manual supervision, the factors in the fire extinguishing experiment report are solidified into the automatic test report output system, and the manual auditing mechanism is supplemented, the distortion problem of the report caused by human errors is avoided, and the performance standardization test of the fire extinguisher is realized.

[0142] Further, according to the fire extinguishing result output module, after the fire extinguishing action is completed, a comprehensive determination is made on a series of result data such as the flame temperature and the smoke concentration, the fire extinguishing effect data is determined, and a complete fire extinguishing test report is automatically output based on the data. The report that does not meet the standard is adjusted and output by the intelligent report output model after the feedback of manual supervision, the factors in the fire extinguishing experiment report are solidified into the automatic test report output system, and the manual auditing mechanism is supplemented, the distortion problem of the report caused by human errors is avoided, and the performance standardization test of the fire extinguisher is realized.

[0143] According to the fire extinguisher performance test method provided in the embodiment of the present application, the target fire field environment can be simulated based on the environmental factors that have an influence on the combustion and the fire extinguishing effect according to the applicability of the target fire extinguisher, and then the fire extinguishing robot performs the fire extinguishing operation on the target fire field environment by using the target fire extinguisher according to the preset control mode of the fire extinguishing robot. In the process of performing the fire extinguishing operation, the related information reflecting the fire extinguishing effect of the target fire extinguisher can be collected in real time, so as to realize the automatic output of the standardized fire extinguisher performance test result, effectively automate the mechanical repetitive labor in the fire extinguishing experiment, and help to improve the accuracy and safety of the fire extinguisher fire extinguishing performance test. Therefore, the problems in the related art that the detection result is greatly influenced by the experience, method, state of the fire extinguishing experiment operator, and factors such as the test environment temperature, humidity, and ventilation degree, which may cause a large performance test error, and the operator needs to frequently approach the fire source, which has a safety risk, are solved.

[0144] Secondly, the fire extinguisher performance test system according to the embodiment of the present application is described with reference to the accompanying drawings.

[0145] Figure 6 is a structural schematic diagram of the fire extinguisher performance test system according to the embodiment of the present application.

[0146] As Figure 6 shown, the fire extinguisher performance test system 10 includes a fire field simulation module 100, a robot control module 200, and a result determination module 300.

[0147] Specifically, the fire field simulation module 100 is configured to simulate a target fire field environment based on environmental factors that have an influence on combustion and extinguishing effect according to applicability of a target fire extinguisher.

[0148] The robot control module 200 is configured to control the fire extinguishing robot to perform the fire extinguishing operation on the target fire field environment by using the target fire extinguisher according to a preset control mode of the fire extinguishing robot.

[0149] The result determination module 300 is configured to collect information reflecting the extinguishing effect of the target fire extinguisher in real time during the fire extinguishing operation, so as to generate a performance test result of the target fire extinguisher.

[0150] It should be noted that the foregoing description of the fire extinguisher performance test method embodiments is also applicable to the fire extinguisher performance test system of the embodiments, which will not be described herein again.

[0151] According to the fire extinguisher performance test system provided in the embodiments of the present application, the target fire field environment can be simulated based on environmental factors that have an influence on combustion and extinguishing effect according to applicability of a target fire extinguisher, and then the fire extinguishing robot can be controlled to perform the fire extinguishing operation on the target fire field environment by using the target fire extinguisher according to a preset control mode of the fire extinguishing robot. In the process of performing the fire extinguishing operation, the information reflecting the extinguishing effect of the target fire extinguisher can be collected in real time, so as to automatically output the standardized fire extinguisher performance test result, effectively automate the mechanical repetitive labor in the fire extinguishing experiment, and help to improve the accuracy and safety of the fire extinguisher performance test.

[0152] The embodiments of the present application also provide a computer program product, which comprises computer programs / instructions, and the computer programs / instructions are executed by a processor to implement the fire extinguisher performance test method as described above.

[0153] In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the present specification and the features of the different embodiments or examples without contradiction.

[0154] Furthermore, the terms "first", "second", "third", etc. are used herein only to describe different steps in a flowchart, and do not imply or suggest a relative importance of the steps so designated. That is, a step designated "first", "second", etc. can implicitly or explicitly include one or more of the steps so designated. In the description of the present application, the meaning of "N" is at least two, such as two, three, etc., unless otherwise explicitly specified.

[0155] Any process or method described in flowcharts or otherwise described herein can be understood as representing a module, segment, or portion of code that includes one or N executable instructions for implementing the specified logical function or process, and the scope of preferred embodiments of the present application encompasses additional implementations that can not be shown or discussed explicitly, including implementations in which the functions are performed in a different order, substantially concurrently, or in reverse order, as appropriate to the functionality involved.

[0156] The logic and / or steps represented in flowcharts or otherwise described herein, for example, can be considered as a list of executable instructions for implementing the logic function, and can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor- containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For purposes of this specification, a "computer-readable medium" can be any apparatus that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electronic connection having one or N wires (electronic apparatus), a portable computer diskette (magnetic apparatus), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber (optical apparatus), and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via the optical scan of the paper or other medium, then compiled, interpreted, or otherwise processed in a suitable manner, if necessary, and then stored in a computer memory.

[0157] It should be understood that portions of the application can be realized with a combination of hardware, software, firmware, or their combination. In the above-described embodiments, the N steps or methods can be realized with software or firmware stored in a memory and executed by a suitable instruction execution system. If realized with hardware and in another embodiment, it can be realized with any one or a combination of the following technologies known in the art: discrete logic circuit with logic gate circuit for implementing logic functions on data signals, application specific integrated circuit with suitable combination logic gate circuit, programmable gate array (PGA), field programmable gate array (FPGA), etc.

[0158] Those skilled in the art of the present technology can understand that all or part of the steps carried out by the above-mentioned embodiments can be completed by a program instructing the relevant hardware, and the program can be stored in a computer readable storage medium. When the program is executed, it includes one of the steps of the method embodiments or a combination thereof.

[0159] In addition, the functional units in each embodiment of the present application can be integrated into one processing module, or each unit can be physically present separately, or two or more units can be integrated into one module. The above-mentioned integrated module can be realized in the form of hardware or in the form of a software function module. The integrated module, if realized in the form of a software function module and sold or used as an independent product, can also be stored in a computer readable storage medium.

[0160] The above-mentioned storage medium can be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it should be understood that the above-mentioned embodiments are exemplary and cannot be understood as limiting the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above-mentioned embodiments within the scope of the present application.

Claims

1. A method of testing the performance of an extinguisher, characterised by, The method comprises the following steps: According to the applicability of the target fire extinguisher, simulate the target fire environment based on environmental factors that have an impact on combustion and fire extinguishing effect; According to the preset control mode of the fire extinguishing robot, control the fire extinguishing robot to perform fire extinguishing operation on the target fire environment with the target fire extinguisher; During the execution of the fire extinguishing operation, real-time collection of relevant information reflecting the fire extinguishing effect of the target fire extinguisher is performed to generate performance test results of the target fire extinguisher.

2. The method of claim 1, wherein, The control mode includes a manual control mode; According to the preset control mode of the fire extinguishing robot, the fire extinguishing robot performs fire extinguishing operation on the target fire environment with the target fire extinguisher, comprising: In response to the preset control mode of the fire extinguishing robot being a manual control mode, according to the manual fire extinguishing operation instruction issued by the operator, the fire extinguishing robot performs fire extinguishing operation on the target fire environment with the target fire extinguisher.

3. The method of claim 1, wherein, The control mode includes an automatic control mode; According to the preset control mode of the fire extinguishing robot, the fire extinguishing robot performs fire extinguishing operation on the target fire environment with the target fire extinguisher, comprising: In response to the preset control mode of the fire extinguishing robot being an automatic control mode, import a standard fire extinguishing operation model matched with the target fire extinguisher; Perform model analysis on the standard fire extinguishing operation model to obtain standard operation information of the fire extinguishing robot; According to the standard operation information of the fire extinguishing robot, control the fire extinguishing robot to perform fire extinguishing operation on the target fire environment with the target fire extinguisher.

4. The method according to any one of claims 1 to 3, characterized in that, The control mode includes an automatic intelligent control mode; According to the preset control mode of the fire extinguishing robot, the fire extinguishing robot performs fire extinguishing operation on the target fire environment with the target fire extinguisher, comprising: In response to the preset control mode of the fire extinguishing robot being an automatic intelligent control mode, collect feedback information after the fire extinguishing robot performs current fire extinguishing operation with the target fire extinguisher; According to the feedback information, adjust the behavior information of the fire extinguishing robot, and perform fire extinguishing operation on the target fire environment based on the adjusted behavior information of the fire extinguishing robot.

5. The method of claim 4, wherein, The feedback information includes environmental state information, combustion state information, fire extinguishing robot state information, fire extinguishing robot limb state information, fire extinguisher state information, and cost generated in the fire extinguishing process; According to the feedback information, adjust the behavior information of the fire extinguishing robot, comprising: Serializing the environmental state information, combustion state information, fire extinguishing robot state information, fire extinguishing robot limb state information, fire extinguisher state information, and cost generated in the fire extinguishing process to obtain an input feature vector; Input the input feature vector into a fire extinguishing robot behavior driving model to predict the behavior information of the fire extinguishing robot at the next moment; Import expert demonstration behavior information from a pre-built expert demonstration data set, and compare the expert demonstration behavior information with the predicted next time behavior information of the fire extinguishing robot, and optimize the parameters of the fire extinguishing robot behavior driving model with the minimum of the pre-built total loss function as the target, and the minimum of the total cost including safety cost, path cost and energy consumption cost as the constraint; Use the parameter-optimized fire extinguishing robot behavior driving model to re-predict the next time behavior information of the fire extinguishing robot, so as to obtain the adjusted behavior information of the fire extinguishing robot; The fire extinguishing robot driving model is constructed based on a multi-head attention mechanism neural network model; the multi-head attention mechanism neural network model includes an encoder and a decoder; the encoder is used to encode the input feature vector to obtain an encoded feature vector; and the decoder is used to decode according to the encoded feature vector and the input feature vector to predict the next time behavior information of the fire extinguishing robot.

6. The method of claim 5, wherein, The expression of the total loss function is as follows: ， wherein, represents the total loss function; describes all internal fire-fighting robot activity parameters; N represents the number of expert demonstration data sets; M represents the number of actual operation data sets of the fire-fighting robot; represents the expert demonstration data set, which contains a number of complete expert demonstration action trajectories; represents the expert demonstration behavior information; represents the predicted acquired behavior information of the fire-fighting robot at the next time; , represents the trajectory loss cost of ; represents the average loss cost of ; represents the fire-fighting experiment state at the kth step; represents the fire-fighting robot action at the kth step; represents the weighted loss cost brought by the state and the fire-fighting robot action ; represents the generation probability of , is the initial state distribution ; is the initial state of the fire-fighting experiment; K represents the total number of trajectory paths in the entire fire-fighting process; represents the probability of selecting the fire-fighting robot action under the state ; represents the probability of transferring to the state after selecting the fire-fighting robot action under the state ; represents the fire-fighting experiment state at the k+1th step; represents the trajectory loss cost of ; represents the relative preference for other trajectories; represents the trajectory negative cost of ; represents the injury state of the fire-fighting robot at the kth step, , , , respectively represent the scalding, choking, and collision of the fire-fighting robot; represents the safety loss weight.

7. The method of claim 5, wherein, The expression of the serialized encoding of the environmental state information, the combustion state information, the fire extinguishing robot state information, the fire extinguishing robot limb state information, the fire extinguisher state information and the cost generated in the fire extinguishing process is as follows: ； wherein, is the fire extinguishing experiment state vector; is an Embedding sequence encoding function; represents a concatenation operation that concatenates all state information into a complete vector; , , , , and are data vectorized representations of the environment state information, the burning state information, the fire extinguishing robot state information, the fire extinguishing robot limb state information, the fire extinguisher state information, and the cost generated by the fire extinguishing process, respectively; represents an offset encoding of the time dimension; wherein ; , , , respectively represent ambient temperature, ambient humidity and ambient smoke concentration; ; , , respectively represent the combustion temperature, the combustion material type and the combustion range; ; , , , , , , , respectively represent visual information, auditory information, tactile information, olfactory information, sensed temperature information, motion information, location information, and remaining consumable energy of the fire-fighting robot; ; body position and orientation, head position and orientation, arm position and orientation, and finger position and orientation of the fire extinguishing robot, respectively; ; respectively represent the type, weight, internal pressure, position, angle of the fire extinguisher, the position of the pull valve, the angle of the pull valve, the discharge rate of the nozzle and the remaining fire extinguisher mass.

8. The method of claim 1, wherein, The related information reflecting the extinguishing effect of the target fire extinguisher at least includes combustion state information and injury degree of the fire extinguishing robot; The performance test result of the target fire extinguisher includes: In response to the injury degree of the fire extinguishing robot exceeding a set threshold, a test performance report of fire extinguishing test failure is generated; In response to the injury degree of the fire extinguishing robot not exceeding a set threshold and the combustion state information satisfying the extinguishing success judgment standard, a test performance report of fire extinguishing test success is generated.

9. A fire extinguisher performance testing system, characterized by, It includes: A fire field simulation module for simulating a target fire field environment based on environmental factors that affect combustion and extinguishing effect according to the applicability of a target fire extinguisher; A robot control module for controlling the fire extinguishing robot to perform fire extinguishing operation on the target fire field environment using the target fire extinguisher according to a pre-set control mode of the fire extinguishing robot; A result determination module for generating a performance test result of the target fire extinguisher by collecting related information reflecting the extinguishing effect of the target fire extinguisher in real time during the fire extinguishing operation.

10. A computer program product comprising computer programs / instructions, characterized in that, The computer program / instructions are executed by the processor to realize the steps of the method of any one of claims 1-8.

Images