Intelligent fire-fighting linkage control method for oil and gas module fire

Abstract

The invention discloses an oil and gas module fire intelligent fire-fighting linkage control method, which comprises the following steps: arranging multiple sensors in an oil and gas module space site, constructing a fire behavior prediction neural network model, training the fire behavior prediction neural network model, outputting a fire occurrence probability, and then carrying out fire extinguishing numerical simulation analysis on carbon dioxide fire extinguishing equipment. And determining the equipment parameters with the optimal fire extinguishing effect under the same fire behavior, generating a fire-fighting linkage decision according to the obtained equipment parameters with the optimal fire extinguishing effect under the same fire behavior, and making a series of collaborative fire-fighting linkage schemes. The fire data of the oil and gas module can be comprehensively and accurately obtained through monitoring of multiple sensors. The fire behavior prediction model constructed based on the neural network can comprehensively analyze historical and real-time data and scientifically predict the fire behavior.

Description

Technical Field

[0001] The present invention relates to a fire fighting method, and more particularly to an intelligent fire fighting linkage control method for an oil and gas module fire. Background Art

[0002] The existing intelligent control method for oil and gas module fire fighting is an intelligent control method based on the linkage of sensor monitoring and an automatic fire extinguishing system. The existing method has the disadvantages of false alarms and missed alarms of sensors and poor automatic fire extinguishing effect. The intelligent discrimination ability of fire sensors in the face of complex fire situations is not strong, and the fire extinguishing system is difficult to effectively cover the fire source in the complex oil and gas module environment, resulting in low fire extinguishing efficiency. Summary of the Invention

[0003] The purpose of the present invention is to overcome the disadvantages of the existing technology and provide an intelligent fire fighting linkage control method for an oil and gas module fire, which greatly improves the accuracy of fire monitoring and early warning, the scientificity of fire extinguishing decision-making, and the effectiveness of fire extinguishing actions.

[0004] To achieve the above purpose, the present invention provides the following technical solutions:

[0005] An intelligent fire fighting linkage control method for an oil and gas module fire of the present invention includes the following steps:

[0006] Step 1: Arrange multiple sensors in the space of the oil and gas module. The multiple sensors include a smoke sensor, a temperature sensor, a flame sensor, a humidity sensor, and a wind speed sensor. The smoke sensor is used to obtain the concentration of smoke and the location where the smoke is generated in the initial stage of a fire; the temperature sensor is used to continuously monitor the temperature in the space of the oil and gas module; the flame sensor is used to receive the infrared ray information radiated by the flame combustion to determine whether a flame is generated and the location of the fire source; the humidity sensor is used to continuously monitor the humidity in the space of the oil and gas module; the wind speed sensor is used to monitor the wind speed at the inlet and outlet of the oil and gas module site;

[0007] Step 2: Determine the architecture of the neural network and construct a fire situation prediction neural network model;

[0008] Step 3: Input the historical data of smoke concentration, temperature, and infrared radiation intensity into the fire situation prediction neural network model, train the fire situation prediction neural network model, and output the probability of a fire occurring;

[0009] Step 4: Input the data of the smoke sensor, temperature sensor, and flame sensor into the fire situation prediction neural network model in real time, and output the probability of a fire occurring When it is determined that no fire has occurred and the fire fighting equipment is not activated. When it is determined that a fire has occurred, step 5 is executed;

[0010] Step Five: Calculate the degree of fire occurrence ε based on the data of smoke concentration, temperature, and infrared radiation intensity. The formula is:

[0011]

[0012] In the formula, C is the real-time smoke concentration output by the smoke sensor, T is the real-time temperature output by the temperature sensor, and I is the real-time infrared radiation intensity output by the flame sensor; C max is the preset maximum smoke concentration value, T max is the preset highest temperature value, and I max is the preset maximum infrared radiation intensity value; w C is the smoke concentration weight, w T is the temperature weight, and w I is the infrared radiation intensity weight;

[0013] Step Six: Use fire dynamics simulation software to conduct numerical simulation analysis of the carbon dioxide fire extinguishing equipment, and determine the equipment parameters with the best fire extinguishing effect under the same fire conditions;

[0014] Step Seven: Generate a fire-fighting linkage decision based on the degree of fire occurrence of the oil and gas module in Step Five and the equipment parameters with the best fire extinguishing effect obtained from the numerical simulation analysis of fire extinguishing in Step Six, and formulate a series of coordinated fire-fighting linkage plans; at the same time, observe the fire situation through video monitoring, and adjust the fire extinguishing equipment parameters according to the actual fire situation to achieve complete fire extinguishing.

[0015] Compared with the prior art, the present invention has the following beneficial effects:

[0016] Through multi-sensor monitoring, the fire situation data of the oil and gas module can be comprehensively and accurately obtained. The fire situation prediction model constructed based on the neural network can comprehensively analyze historical and real-time data and scientifically predict the fire situation. Fire-fighting linkage decisions are generated according to different degrees of fire occurrence, and real-time observation and optimization can be carried out, greatly improving the accuracy of fire monitoring and early warning, the scientificity of fire-fighting decisions, and the effectiveness of fire-fighting operations. Brief Description of the Drawings

[0017] Figure 1 It is a flowchart of an intelligent fire-fighting linkage control method for an oil and gas module according to an embodiment of the present invention. Detailed Embodiment

[0018] The present invention will be described in detail below in conjunction with specific embodiments and the drawings:

[0019] As Figure 1 shown, an intelligent fire-fighting linkage control method for an oil and gas module includes the following steps:

[0020] Step 1: Arrange multiple sensors in the space of the oil and gas module. The multiple sensors include a smoke sensor, a temperature sensor, a flame sensor, a humidity sensor, and a wind speed sensor. The smoke sensor is used to obtain the concentration of smoke and the location where the smoke is generated at the initial stage of a fire; the temperature sensor is used to monitor the temperature in the space of the oil and gas module in real time; the flame sensor is used to receive the infrared ray information radiated by the flame combustion to determine whether a flame is generated and the location of the fire source; the humidity sensor is used to monitor the humidity in the space of the oil and gas module in real time; the wind speed sensor is used to monitor the wind speed at the inlet and outlet of the oil and gas module site.

[0021] Step 2: Determine the architecture of the neural network and construct a fire prediction neural network model. The specific steps can be as follows:

[0022] The architecture of the neural network includes an input layer, a hidden layer, and an output layer.

[0023] The input layer is used to receive the concentration of smoke output by the smoke sensor, the temperature output by the temperature sensor, and the infrared radiation intensity output by the flame sensor.

[0024] The hidden layer is arranged between the input layer and the output layer and contains five neurons. These neurons perform a weighted summation operation on the input data and perform a non-linear transformation through the Sigmoid activation function, so as to effectively extract the key features in the data.

[0025] The Sigmoid activation function is as follows:

[0026]

[0027] In the formula, x is the result after the neuron performs a weighted summation operation on the input data.

[0028] The output layer consists of one neuron, which outputs the probability value of a fire occurring. This neuron is trained to convert the data processed by the hidden layer into a value between 0 and 1.

[0029] The output of the i-th neuron in the single hidden layer network is:

[0030]

[0031] In the formula, ω hj is the connection weight from the h-th hidden layer neuron to the j-th output, ψ j is the activation function of the h-th hidden layer neuron, ν ih is the connection weight from the i-th input to the h-th hidden layer neuron, b h is the threshold of the h-th hidden layer neuron, q is the number of hidden layer neurons, and d is the number of input layer neurons.

[0032] Step 3: Input the historical data of smoke concentration, temperature, and infrared radiation intensity into the fire prediction neural network model, train the fire prediction neural network model, and output the probability of a fire occurring.

[0033] In this step, the fire prediction neural network model can automatically identify the relationship between the data of each sensor and the occurrence of a fire through learning a large amount of historical data. The training process of the fire prediction neural network model can be as follows:

[0034] First step: Perform data preprocessing on the historical data of each fire sensor.

[0035] Perform data cleaning operations on the historical data of each fire sensor, including the concentration of smoke, temperature, and infrared radiation intensity, remove outliers and noise data and other factors that may interfere with subsequent analysis, and use the method of normalization processing to map the data to the interval of 0 - 1. The normalization formula is:

[0036]

[0037] In the formula, x is the original data, x min is the minimum value of the data, x max is the maximum value of the data, x new is the normalized data.

[0038] Second step: Input the preprocessed data into the fire prediction neural network model for model training.

[0039] Third step: Use the mean square error (MSE) as the loss function, that is:

[0040]

[0041] where n represents the number of samples, y i represents the true probability of a fire occurring, represents the probability of a fire occurring predicted by the network.

[0042] Fourth step: Use the backpropagation algorithm to dynamically adjust the weights of the fire prediction neural network model until the loss function is minimized to obtain the final fire prediction neural network model. This model can calculate the probability of a fire occurring based on real-time fire sensor data.

[0043] Step 4: Input the data of the smoke sensor, temperature sensor, and flame sensor into the fire prediction neural network model in real time, and output the probability of a fire occurring When it is determined that no fire has occurred, and the fire-fighting equipment is not activated. When it is determined that a fire has occurred, and step 5 is executed;

[0044] Step 5: Calculate the degree of fire occurrence ε based on the data of smoke concentration, temperature, and infrared radiation intensity. The formula is:

[0045]

[0046] In the formula, C is the real-time smoke concentration output by the smoke sensor, T is the real-time temperature output by the temperature sensor, and I is the real-time infrared radiation intensity output by the flame sensor. C max is the preset maximum smoke concentration value, T max is the preset maximum temperature value, and I max is the preset maximum infrared radiation intensity value. w C is the weight of smoke concentration, w T is the weight of temperature, and w I is the weight of infrared radiation intensity.

[0047] Determine the weight coefficients of smoke concentration, temperature, and infrared radiation intensity through the Delphi method. The specific process is as follows: Form an expert group from the field of fire protection in the oil and gas module, conduct multiple rounds of anonymous questionnaires, summarize and feedback the average weight results to the experts for adjusting the scores after each round until the weight distribution tends to be stable and reaches a consensus, and determine the final weight coefficients.

[0048] Step 6: Use fire dynamics simulation software to conduct numerical simulation analysis of the carbon dioxide fire extinguishing equipment, and determine the equipment parameters with the best fire extinguishing effect under the same fire conditions. The specific steps are as follows:

[0049] First step: According to the size and shape of the actual oil and gas module site, including the layout of oil tanks, pipelines, equipment, etc., construct a three-dimensional geometric model of the oil and gas module site in the pre-processing module of the software.

[0050] Second step: According to the physical properties of the oil and gas products and storage and transportation equipment in the oil and gas module site, set the combustion characteristics of the oil and gas products (such as oil products, natural gas), such as parameters like heat of combustion, flash point, ignition point, etc.; set the thermophysical properties of the storage and transportation equipment (such as oil tanks, pipelines), such as thermal conductivity, specific heat capacity, etc.; define the physical properties of carbon dioxide, such as density, specific heat capacity, etc.

[0051] Third step: According to the temperature data monitored by the temperature sensor, the infrared radiation intensity data monitored by the flame sensor, the humidity data monitored by the humidity sensor, and the wind speed data monitored by the wind speed sensor in the oil and gas module site, set the environmental parameters of the oil and gas module site in the software: such as environmental temperature, humidity, and wind speed conditions, set the parameters of the fire source: such as location, size, heat release rate, etc. parameters, and set the inlet and outlet boundary conditions and wall types of the oil and gas module site.

[0052] Step 4: Set the parameters of the carbon dioxide fire extinguishing equipment, including the position of the nozzles, the spraying angle, the flow coefficient, and the carbon dioxide release time curve.

[0053] Step 5: Start the simulation calculation. The software solves to obtain the numerical solutions of the flame extinction time, the oxygen concentration reduction rate, the temperature drop rate, and the carbon dioxide distribution varying with time during the fire development process.

[0054] Step 6: According to the simulation results, evaluate the fire extinguishing effects under different equipment parameters and the same fire situation, and obtain the equipment parameters with the best fire extinguishing effect: determine the best positions of the nozzles, the spraying angles, the flow coefficients, and the carbon dioxide release time curve of the carbon dioxide equipment by observing the flame extinction time, the oxygen concentration reduction rate, and the temperature drop rate calculated by the software simulation under different fire extinguishing equipment parameters.

[0055] Step 7: Generate a fire-fighting linkage decision based on the degree of the oil and gas module fire in Step 5 and the equipment parameters with the best fire extinguishing effect obtained from the fire extinguishing numerical simulation analysis in Step 6 under the same fire situation, and formulate a series of coordinated fire-fighting linkage plans to ensure effective fire extinguishing of the oil and gas module; at the same time, observe the fire situation through video monitoring, and adjust the fire extinguishing equipment parameters according to the actual fire situation for complete fire extinguishing.

[0056] Among them, the specific steps for formulating a series of coordinated fire-fighting linkage plans are as follows:

[0057] When it is determined that ε≤0.3, it indicates that the degree of the fire is small. The computer outputs control signals to the audible and visual alarm and the voice broadcast. The audible and visual alarm warns people with high-intensity and high-frequency signals. At the same time, the voice broadcast will clearly inform the on-site personnel of the specific location and the degree of danger of the oil and gas module fire. The relevant personnel quickly make a fire extinguishing response, and at this time, the fire-fighting linkage system is not triggered;

[0058] When 0.3<ε≤0.7, the degree of the fire is moderate. The computer outputs control signals to the audible and visual alarm, the voice broadcast, and the fire-fighting equipment. The audible and visual alarm warns people with high-intensity and high-frequency signals. At the same time, the voice broadcast will clearly inform the on-site personnel of the specific location and the degree of danger of the oil and gas module fire. At this time, according to the real-time fire situation data and the fire extinguishing numerical simulation analysis results, select the fire extinguishing strategy of the carbon dioxide equipment with the best fire extinguishing effect for fire extinguishing;

[0059] When ε>0.7, the degree of the fire is large. The computer outputs control signals to the audible and visual alarm, the fire-fighting equipment, the emergency lighting system, and the evacuation indicator lights. The audible and visual alarm warns people with high-intensity and high-frequency signals, triggers the high-pressure fire extinguishing strategy of the carbon dioxide equipment and starts the emergency lighting system and the evacuation indicator lights. The staff orderly guides the evacuation according to the personnel distribution in different areas, closes the fire-fighting facilities such as the fire door curtains, and conducts fire extinguishing.

[0060] Observe the fire situation through video monitoring and real-time fire data. Video monitoring can provide intuitive images of the fire scene. If obvious flames can still be observed after the flame extinction time obtained from the fire extinguishing numerical simulation arrives, increase the injection pressure and release amount of the carbon dioxide fire extinguishing equipment and extend the working time of the fire extinguishing equipment to ensure that the flames are completely extinguished.

Claims

1. An intelligent fire fighting linkage control method for oil and gas module fire, characterized in that The following steps are involved: Step 1: multiple sensors are arranged in the oil and gas module space, wherein the multiple sensors include a smoke sensor, a temperature sensor, a flame sensor, a humidity sensor and a wind speed sensor. The smoke sensor is used to obtain the concentration of smoke and the location of smoke generation at the initial stage of a fire; the temperature sensor is used to monitor the temperature in the oil and gas module space in real time; the flame sensor is used to receive infrared information radiated by flame combustion to determine whether a flame is generated and the location of the fire source; the humidity sensor is used to monitor the humidity in the oil and gas module space in real time; the wind speed sensor is used to monitor the wind speed at the inlet and outlet of the oil and gas module space; Step 2: Determine the architecture of the neural network and build a fire prediction neural network model; Step 3: Input historical data of smoke concentration, temperature, and infrared radiation intensity into the fire prediction neural network model, train the fire prediction neural network model and output the probability of fire occurrence; Step 4: Input the data of smoke sensor, temperature sensor and flame sensor into the fire prediction neural network model in real time and output the probability of fire occurrence. when When it is determined that no fire has occurred, the fire fighting equipment will not be activated. If it is determined that a fire has occurred, execute step five; Step 5: Calculate the fire occurrence degree ε based on the smoke concentration, temperature, and infrared radiation intensity data. The formula is: Where C is the real-time smoke concentration output by the smoke sensor, T is the real-time temperature output by the temperature sensor, and I is the infrared radiation intensity output by the flame sensor in real time; C max is the preset maximum smoke density value, T max is the preset maximum temperature value, I max is the preset maximum infrared radiation intensity value; w C is the smoke density weight, w T is the temperature weight, w I is the infrared radiation intensity weight; Step 6: Use fire dynamics simulation software to perform numerical simulation analysis of carbon dioxide fire extinguishing equipment to determine the equipment parameters with the best fire extinguishing effect under the same fire conditions; Step 7. Generate fire linkage decisions based on the extent of the oil and gas module fire in step 5 and the equipment parameters with the best fire extinguishing effect under the same fire conditions obtained from the fire extinguishing numerical simulation analysis in step 6, and formulate a series of coordinated fire linkage plans; at the same time, observe the fire situation through video surveillance, and adjust the fire extinguishing equipment parameters according to the actual fire situation to completely extinguish the fire.

2. The intelligent fire fighting linkage control method for oil and gas module fire according to claim 1 is characterized in that: The specific steps of step 2 are as follows: The architecture of the neural network includes an input layer, a hidden layer and an output layer; The input layer is used to receive the smoke concentration output by the smoke sensor, the temperature output by the temperature sensor, and the infrared radiation intensity output by the flame sensor; The hidden layer is arranged between the input layer and the output layer, and contains five neurons. The five neurons perform a weighted sum operation on the input data and perform a nonlinear transformation through a Sigmoid activation function to extract key features from the data. The Sigmoid activation function is: Where x is the result of the neuron performing a weighted sum operation on the input data; The output layer consists of one neuron, which outputs the probability value of fire occurrence; The neuron is trained to convert the data processed by the hidden layer into a value between 0 and 1; The output of the i-th neuron in a single hidden layer network is for: Where ω hj is the connection weight from the hth hidden layer neuron to the jth output, ψ j is the activation function of the hth hidden layer neuron, v ih is the connection weight from the ith input to the hth hidden layer neuron, b h is the threshold of the hth hidden layer neuron, q is the number of hidden layer neurons, and d is the number of input layer neurons.

3. The intelligent fire fighting linkage control method for oil and gas module fire according to claim 1 or 2, characterized in that: The weight coefficients of smoke concentration, temperature and infrared radiation intensity were determined by the Delphi method.

4. The intelligent fire fighting linkage control method for oil and gas module fire according to claim 1 or 2, characterized in that: The specific steps of step six are: The first step is to build a three-dimensional geometric model of the oil and gas module site in the pre-processing module of the software; The second step is to set the combustion characteristics of the oil and gas products according to the physical properties of the oil and gas products and storage and transportation equipment in the oil and gas module site; set the thermal physical properties of the storage and transportation equipment; and define the physical properties of carbon dioxide; Step 3: According to the temperature data monitored by the temperature sensor of the oil and gas module site, the infrared radiation intensity data monitored by the flame sensor, the humidity data monitored by the humidity sensor, and the wind speed data monitored by the wind speed sensor, the environmental parameters of the oil and gas module site, the parameters of the fire source, the inlet and outlet boundary conditions, and the wall type of the oil and gas module site are set in the software; Step 4: Set the parameters of the carbon dioxide fire extinguishing equipment; Step 5: Start the simulation calculation, and the software will obtain the numerical solution of the flame extinction time, oxygen concentration reduction rate, temperature drop rate and carbon dioxide distribution change over time during the fire development process; Step 6. Based on the simulation results, evaluate the fire extinguishing effect under different equipment parameters and the same fire conditions, and obtain the equipment parameters with the best fire extinguishing effect: By observing the software simulation and calculating the flame extinguishing time, oxygen concentration reduction rate, and temperature drop rate under different fire extinguishing equipment parameters, determine the best carbon dioxide equipment nozzle position, injection angle, flow coefficient and carbon dioxide release time curve.

5. The intelligent fire fighting linkage control method for oil and gas module fire according to claim 4 is characterized in that: The specific steps for developing a series of coordinated fire protection linkage plans in step 7 are as follows: When it is determined that ε≤0.3, it indicates that the fire is of a minor severity, and the computer outputs control signals to the sound and light alarm and voice broadcast; the sound and light alarm warns personnel with high-intensity, high-frequency signals, and at the same time, the voice broadcast informs the on-site personnel of the specific location and degree of danger of the oil and gas module fire, and the relevant personnel respond to the fire, and the fire linkage system is not triggered at this time; When 0.3<ε≤0.7, the fire is moderate, and the computer outputs control signals to the sound and light alarm, voice broadcast and fire fighting equipment; the sound and light alarm warns personnel with high-intensity and high-frequency signals, and at the same time, the voice broadcast informs the on-site personnel of the specific location and danger level of the oil and gas module fire. At this time, according to the real-time fire data and the results of the fire extinguishing numerical simulation analysis, the carbon dioxide equipment fire extinguishing strategy with the best fire extinguishing effect is selected to extinguish the fire; When ε>0.7, the fire is serious, and the computer outputs control signals to the sound and light alarm, fire-fighting equipment, emergency lighting system and evacuation indicator light; the sound and light alarm warns personnel with high-intensity and high-frequency signals, triggers the high-pressure fire-fighting strategy of the carbon dioxide equipment, and starts the emergency lighting system and evacuation indicator light. The staff guides the evacuation in an orderly manner according to the distribution of personnel in different areas, closes the fire curtain and other fire-fighting facilities, and extinguishes the fire; The fire situation is observed through video surveillance and real-time fire data. Video surveillance provides intuitive images of the fire scene. If obvious flames can still be observed after the flame extinguishing time obtained by the fire extinguishing numerical simulation is reached, the injection pressure and release volume of the carbon dioxide fire extinguishing equipment are increased and the working time of the fire extinguishing equipment is extended to ensure that the flame is completely extinguished.

Images