An adaptive valley oil diffusion blocking device and method

Abstract

The application discloses a self-adaptive valley oil diffusion blocking device and method, which comprises a V-shaped opening shell, upper and lower telescopic mechanisms and parallel plate groups are arranged on the two V-shaped sides of the V-shaped opening shell, the parallel plate groups comprise a plurality of parallel plates which are arranged in an overlapped mode and are connected through hinges, the telescopic ends of the upper and lower telescopic mechanisms are connected with one side of the parallel plates, a microcontroller, a storage battery and a motor for driving the upper and lower telescopic mechanisms are arranged on the V-shaped opening shell, and the storage battery and the motor are electrically connected with the microcontroller; oil leakage detection sensors and density sensors are arranged on the parallel plates; and a pressure sensor and a flow sensor are arranged at the bottom of the V-shaped opening shell. The microcontroller in the device is provided with an advanced algorithm, can intelligently receive and process sensor data, and controls the unfolding and folding of the parallel plates according to the leakage condition. In addition, the real-time transmission of the device state and performance is realized through remote monitoring, and the intelligentization and remote management capability of the system are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to a self-adaptive valley oil diffusion blocking method, belonging to the technical field of oil diffusion blocking. BACKGROUND

[0002] In today's society, with the continuous advancement of industrialization and modernization, chemicals and petroleum products have been widely used in various fields. However, the use of these chemicals and petroleum products also brings potential environmental risks, among which oil leakage incidents are particularly concerned. Oil leakage not only causes serious pollution to land and water sources, but also has long-term effects on the ecosystem and human health. Therefore, developing efficient oil leakage protection technology has become a top priority.

[0003] Currently, although there are some oil leakage response technologies, in special terrain, especially in valley areas, there are a series of challenges. Traditional leakage safety technology is usually difficult to adapt to the complexity of valley terrain, the rapid diffusion of leakage and the irregularity of the height of the leakage site increase the difficulty of responding to challenges. In addition, many existing technologies often have monitoring blind spots, long response time, insufficient adaptability and other problems in actual application, which are difficult to meet the needs of rapid and accurate response to oil leakage in valley areas.

[0004] Therefore, the present application aims to combine the urgency of current leakage safety and the limitations of existing technologies, and provide an innovative oil diffusion blocking device and method suitable for valley terrain, that is, a valley oil diffusion blocking device and method with adaptability, intelligent control and high efficiency. SUMMARY

[0005] In order to overcome the defects in the prior art, the present application aims to provide a self-adaptive valley oil diffusion blocking device and method.

[0006] The technical solution provided by the present application to solve the above technical problems is: a self-adaptive valley oil diffusion blocking device, comprising a V-shaped opening shell, both V-shaped sides of the V-shaped opening shell are provided with an up-down telescopic mechanism and a parallel plate group, the parallel plate group comprises a plurality of parallel plates arranged in overlap, the parallel plates are connected by hinges, the telescopic end of the up-down telescopic mechanism is connected with one side of the parallel plate, the V-shaped opening shell is provided with a microcontroller, a battery, and a motor driving the up-down telescopic mechanism, the battery and the motor are electrically connected with the microcontroller; the parallel plate is provided with an oil leakage detection sensor and a density sensor; the bottom of the V-shaped opening shell is provided with a pressure sensor and a flow sensor.

[0007] Further technical solutions are that the top surface of the V-shaped opening shell is provided with a solar cell panel, and the solar cell panel is electrically connected with the battery.

[0008] Further, the V-shaped opening shell is provided with an oil collecting pipe.

[0009] Further, the V-shaped opening shell is provided with a PVD coating on the outer surface and a titanium dioxide coating on the inner surface.

[0010] An adaptive valley oil diffusion blocking method, comprising the following steps:

[0011] Step 1: Obtain the real-time oil leakage situation in the valley through each sensor;

[0012] Step 2: Analyze the data collected by each sensor, and pre-process the sensor data based on the exponential weighted moving average filtering method;

[0013] Step 3: Based on the fuzzy logic control algorithm, the parameters collected by the sensor are used as input variables, and a fuzzy set is defined for each input variable. Then, a membership function and a fuzzy rule are defined for each fuzzy set. Fuzzy inference is performed using the fuzzy rule and the fuzzy input data, and the fuzzy output value is calculated. The fuzzy output value is converted into a specific estimated leakage amount;

[0014] Step 4: According to the estimated leakage amount, the microcontroller controls the motor and the telescopic mechanism to complete the expansion and contraction of the parallel plates, so as to maximize the interception of the leaked oil;

[0015] Step 5: Determine the wear and service life of the device based on the recurrent neural network predictive maintenance algorithm.

[0016] Further, the specific process of step 2 is as follows:

[0017] Step 21: The original data of oil density, oil pressure, oil infiltration depth, and oil flow rate are used as inputs, each parameter including multiple data points X(i,t). The smoothing parameter a is also used as an input to control the smoothing degree. The smoothed oil parameter data sequence S(i,t) is used as an output;

[0018] Step 22: Initialize the smoothing parameter and the smoothing value, and set the first smoothing value S(i,1) equal to the value of the first data point X(i,1);

[0019] Step 23: Calculate the EWMA value of each sensor until all time steps are processed;

[0020] Step 24: The final output is the smoothed oil parameter data sequence.

[0021] Further, the calculation formula in step 23 is as follows:

[0022] S(i,t) = a * X(i,t) + (1-a) * S(i,t-1)

[0023] In the formula, X(i,t) is a data point, a is a smoothing parameter, and S(i,t) is a smoothed oil parameter data sequence.

[0024] Further, the specific process of the step 3 is:

[0025] Step 31, define fuzzy input: oil density p, infiltration depth D, and fuzzy output: estimated leakage flow Q.

[0026] Step 32, map the oil density p and the infiltration depth D to the fuzzy set respectively, and define the fuzzy membership function and the membership degree.

[0027] Step 33, define the fuzzy rule in the form of if-then: each rule includes a condition part and a conclusion part.

[0028] Step 34, the fuzzy inference uses the fuzzy rule and the fuzzified input data to perform fuzzy inference, and calculates the fuzzy output value of the estimated leakage.

[0029] Step 35, the specific estimated leakage after defuzzification is obtained by calculating the barycentric position on the fuzzy output membership degree curve.

[0030] Further, the specific process of the step 5 is:

[0031] Step 51, collect data related to the device, pre-process the data after screening and conversion, and then divide the data set into a training set and a test set.

[0032] Step 52, select the long short-term memory network and the gated recurrent unit in the recurrent neural network architecture to capture the long-term dependence relationship in the sequence data.

[0033] Step 53, build an RNN model, and define the input layer, hidden layer and output layer of the RNN.

[0034] Step 54, input the training set into the RNN model, perform back propagation and weight update to adjust the model parameters, and obtain the trained RNN model.

[0035] The present application has the following beneficial effects:

[0036] 1. In the device design, through the V-shaped opening device and the adjustable parallel plate, the device can quickly respond to different leakage conditions, realize self-adaptive interception of the leaked oil in the valley, improve the interception efficiency and prevent the oil from spreading to the maximum extent.

[0037] 2. Solar energy driving and energy storage design: solar panels convert sunlight into electricity, which is stored in batteries to ensure stable operation of the device even in the absence of sunlight or at night; PVD coating is added to the outer surface of the device shell to improve its wear resistance and corrosion resistance; the inner surface is coated with titanium dioxide, which can utilize photocatalytic function to decompose the accumulated dirt inside the shell caused by leaked oil, achieving self-cleaning function;

[0038] 3. The microcontroller in the device is equipped with advanced algorithms that can intelligently receive and process sensor data, and control the expansion and contraction of the parallel plates according to the leakage situation. In addition, it has self-learning and adaptive algorithms, which can realize real-time transmission of device status and performance through remote monitoring, improving the intelligence and remote management ability of the system;

[0039] 4. Predictive maintenance during monitoring to enable the device to predict possible wear and tear based on historical data and environmental changes, helping to maintain and maintain in a timely manner, reducing the probability of sudden failure, and improving the reliability and service life of the device. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 The device structure diagram of the present application;

[0041] Figure 2 The fuzzy logic control process diagram of the present application;

[0042] Figure 3 The predictive maintenance flowchart of the present application. DETAILED DESCRIPTION

[0043] The technical solutions of the present application will be described in detail below in conjunction with the drawings. Obviously, the described embodiments are part of the embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor belong to the scope of protection of the present application.

[0044] As Figure 1 shown, the adaptive valley oil diffusion blocking device of the present application comprises a V-shaped opening shell 1, which presents a V shape from the center to both ends, and the two sides gradually extend outward;

[0045] Both V-shaped sides of the V-shaped open shell 1 are provided with up-down telescopic mechanisms 5 and parallel plate groups 6, the up-down telescopic mechanisms 5 are located on the lower side of the parallel plate groups 6, the parallel plate groups 6 include a plurality of overlapping parallel plates, the up-down parallel plates are connected through hinges, the telescopic ends of the up-down telescopic mechanisms 5 are connected with one side of the parallel plates, the V-shaped open shell 1 is provided with a microcontroller 4, a storage battery 2 and a motor 3 for driving the up-down telescopic mechanisms 5, the storage battery 2 and the motor 3 are electrically connected with the microcontroller 4; the parallel plates are provided with oil leakage detection sensors and density sensors for detecting the existence of oil leakage; the bottom of the V-shaped open shell 1 is provided with a pressure sensor and a flow sensor for measuring the flow and infiltration depth of the leaked oil flowing through the device; the top surface of the V-shaped open shell 1 is provided with a solar panel, the solar panel is electrically connected with the storage battery 2, and the solar energy is converted into electrical energy through a solar energy converter to charge the storage battery 2.

[0046] The microcontroller 4 is configured with advanced algorithms for receiving and processing sensor data, automatically optimizing the deployment and folding of the parallel plates based on historical sensor data records, and improving interception efficiency. The microcontroller 4 is also configured with a wireless communication module with remote monitoring function, which can send status and performance data to a remote server for remote monitoring;

[0047] In the design of the barrier device, an oil collection pipe 7 and an oil collection net are provided, which can extract and store the intercepted oil for subsequent further processing or recycling. This helps to minimize the impact on the environment and promotes the sustainable management and utilization of oil.

[0048] In the implementation process, the V-shaped open shell 1 is made of stainless steel material, the surface is coated with PVD coating to improve the wear resistance and corrosion resistance of the metal, to ensure reliable operation in harsh environments, and the inner surface is coated with titanium dioxide coating, which can utilize the photocatalytic function to decompose the accumulated dirt of the leaked oil inside the shell, realizing the self-cleaning function.

[0049] A valley oil diffusion barrier method, the above barrier device is set as three barriers with an interval of 35 meters and 50 meters according to the leakage direction of the oil along the hillside, and independently completes monitoring, blocking and information feedback, the steps are as follows:

[0050] Step 1, obtain the real-time leakage situation of the oil in the valley through each sensor, including pressure, flow and infiltration depth;

[0051] Step 2, analyze the data collected by each sensor, and pre-process the sensor data based on the exponential weighted moving average filtering method; through the filtering algorithm, the sensor data is smoothed, and the new data point is given a greater weight to adapt to changes, so as to eliminate unstable data factors;

[0052] Step 21, the original data of oil density, oil pressure, oil infiltration depth, oil flow rate are taken as input, each parameter includes multiple data points X(i,t); smoothing parameter a is also taken as input, used to control the degree of smoothing; smoothed oil parameter data sequence S(i,t) is taken as output;

[0053] Step 22, initialize smoothing parameter and smoothing value, make the first smoothing value S(i,1) equal to the value of the first data point X(i,1);

[0054] Step 23, calculate EWMA value according to the following formula; repeat the calculation for each sensor data until all time steps are processed;

[0055] S(i,t) = a × X(i,t) + (1-a) × S(i,t-1)

[0056] In the formula: X(i,t) is a data point; a is a smoothing parameter; S(i,t) is a smoothed oil parameter data sequence

[0057] Step 24, the final output is the smoothed oil parameter data sequence, where the value of each time step is an exponentially weighted moving average value.

[0058] Step 3, based on the algorithm of fuzzy logic control, take the parameters collected by the sensor as input variables, define fuzzy sets for each input variable, define membership functions and fuzzy rules for each fuzzy set, use fuzzy rules and fuzzy input data to perform fuzzy reasoning, calculate fuzzy output value, and convert fuzzy output value to specific estimated leakage;

[0059] Step 31, define fuzzy input: oil density p, infiltration depth D, fuzzy output: estimated leakage Q;

[0060] Step 32, map oil density p and infiltration depth D to fuzzy sets (Low, Medium, High) respectively, define fuzzy membership functions and membership degrees;

[0061] Fuzzy set of oil density: Low, Medium, High

[0062] Fuzzy set of oil infiltration depth: Shallow, MediumDepth, Deep

[0063] Step 33, define fuzzy rules in the form of if-then: each rule includes a condition part (if part) and a conclusion part (then part).

[0064] If the oil density is High and the oil penetration depth is Shallow, then the estimated leakage is High.

[0065] If the oil density is Low and the oil penetration depth is Deep, then the estimated leakage is Low;

[0066] Step 34, fuzzy reasoning uses fuzzy rules and fuzzified input data to perform fuzzy reasoning, calculating the fuzzy output value of the estimated leakage;

[0067] It is achieved through fuzzy logic operations, using the min-max method:

[0068] For each fuzzy rule, calculate the minimum value of the rule weight (RW) and the membership of the output part (OM), which is the membership of the fuzzy output value;

[0069] OM = min(RW, OM)

[0070] Step 35, defuzzification, convert the fuzzy output to a specific numerical value to represent the estimated value of the leakage, using the defuzzification method (center of gravity method);

[0071] By calculating the center of gravity position on the fuzzy output membership curve, the specific estimated leakage after defuzzification is obtained;

[0072] Q = ∫(x*F(x))dx / ∫F(x)dx

[0073] In the formula: Q is the output leakage estimate, x is the variable on the fuzzy output membership curve, and F(x) is the fuzzy output membership;

[0074] Step 4, according to the estimated leakage, control the electric motor and telescopic mechanism through the microcontroller to complete the unfolding and folding of the parallel plate, so as to intercept the leaked oil to the greatest extent;

[0075] Step 5, determine the wear and service life of the device based on the recurrent neural network predictive maintenance algorithm;

[0076] Step 51, collect data related to the device, including sensor readings, device operating status, maintenance records, environmental factors, etc. After filtering and converting the data, pre-process the data including cleaning outliers, handling missing data, normalizing or standardizing, etc.

[0077] Divide the data set into training set and test set; usually use time series division to ensure the generalization ability of the model to future data;

[0078] Step 52, select long short-term memory network (LSTM) and gated recurrent unit (GRU) in recurrent neural network (RNN) architecture to capture long-term dependencies in sequence data;

[0079] Step 53, build RNN model, define input layer, hidden layer and output layer of RNN;

[0080] Select mean square error measure model to measure the difference between the output and the true value:

[0081]

[0082] In the formula: y pred,i is the predicted value of the neural network, y ture,i is the true value, and n is the number of samples;

[0083] Select optimizer Adam to update model parameters and minimize loss, and the optimizer updates parameter formula:

[0084]

[0085] In the formula: θ new is the updated parameter, θ old is the old parameter, α is the learning rate, is the gradient of the loss function with respect to the parameter;

[0086] Step 54, model training, input the training set into the RNN model, perform back propagation and weight update to adjust the model parameters, use the validation set to monitor the model performance and avoid overfitting:

[0087] h t = σ(W hh ·h t-1 +W ih ·x t +b h )

[0088] y t = σ(W ho ·h t +b o )

[0089] In the formula: h t is the hidden state of the current time step, x t is the input of the current time step, W hh , W ih , W ho are the weight matrices of the input layer to the hidden layer, the hidden layer to the hidden layer, and the hidden layer to the output layer, respectively, b h , b o are the bias vectors of the hidden layer and the output layer, respectively, and σ is the activation function ReLU;

[0090] Backpropagation (gradient computation):

[0091]

[0092] Step 55, evaluate the performance of the model using the test set and its performance on new data;

[0093] Adjust the hyperparameters such as learning rate, number of hidden layer nodes, etc. according to the performance of the model to improve the robustness of the model. The trained and improved RNN model is updated to the actual device by month to carry out the predictive maintenance of the leakage amount. The RNN model is configured to be integrated with the remote monitoring system to realize real-time feedback and adjustment to ensure the matching degree of the model and the actual operation environment.

[0094] The above description is not any form of limitation on the present application, although the present application has been disclosed by the above examples, however, it is not intended to limit the present application, any person skilled in the art, without departing from the technical solution of the present application, can make some changes or modifications to the equivalent examples of equivalent changes by using the above disclosed technical content, but any simple modification, equivalent change and modification made to the above examples according to the technical essence of the present application, all still belong to the scope of the technical solution of the present application.

Claims

1. An adaptive valley oil diffusion barrier, comprising: The V-shaped open shell (1) is provided with a solar cell panel on the top surface, and the solar cell panel is electrically connected with the battery (2).

2. An adaptive valley oil diffusion barrier according to claim 1, wherein, The V-shaped open shell (1) is provided with an oil collecting pipe (7).

3. An adaptive valley oil diffusion barrier according to claim 1, wherein, The outer surface of the V-shaped open shell (1) is provided with a PVD coating, and the inner surface is provided with a titanium dioxide coating.

4. An adaptive valley oil diffusion barrier according to claim 1, wherein, The method adopts the self-adaptive valley oil diffusion blocking device of any one of claims 1-4 for diffusion blocking, comprising the following steps:

5. A self-adapting method for blocking the diffusion of products in a valley, characterized in that, Step 1, obtain the real-time leakage condition of the oil in the valley through each sensor; Step 2, analyze the data collected by each sensor, and pre-process the sensor data based on the exponential weighted moving average filtering method; Step 3, based on the fuzzy logic control algorithm, the parameters collected by the sensor are used as input variables, and a fuzzy set is defined for each input variable, then a membership function and a fuzzy rule are defined for each fuzzy set, fuzzy inference is performed using the fuzzy rule and the fuzzy input data, the fuzzy output value is calculated, and the fuzzy output value is converted into a specific estimated leakage amount; Step 4, according to the estimated leakage amount, the motor and the telescopic mechanism are controlled by the microcontroller to complete the expansion and contraction of the parallel plates, so as to maximize the interception of the leaked oil; Step 5, the recursive neural network predictive maintenance algorithm is used to determine the wear and service life of the device. The specific process of step 2 is:

6. An adaptive valley hydrocarbon diffusion barrier according to claim 5, wherein, Step 23, calculate the EWMA value of each sensor until all time steps are processed; Step 21, raw data of density, oil product pressure, oil product penetration depth, oil product flow rate of oil product as input, each parameter including multiple data points ; smoothing parameter a , also as input, for controlling the degree of smoothing; smoothed oil product parameter data sequence as output; Step 22, initialize smoothing parameters and smoothing value, make first smoothing value equal to value of first data point ;​ Step 24, the final output is the smoothed oil parameter data sequence. The calculation formula in step 23 is:

7. An adaptive valley hydrocarbon diffusion barrier according to claim 6, wherein, The specific process of step 3 is: wherein: is the data point; a is the smoothing parameter; is the smoothed oil product parameter data sequence.

8. The adaptive valley hydrocarbon diffusion barrier of claim 5, wherein, ρ Step 31, define fuzzy input: oil density ρ , penetration depth D , fuzzy output: estimated leak rate Q ; Step 32, the density of the oil product Step 33, define the fuzzy rule in the form of if-then: each rule includes a condition part and a conclusion part; , the penetration depth D are mapped onto fuzzy sets respectively, and the fuzzy membership functions and the membership degrees are defined; Step 34, fuzzy inference uses fuzzy rules and fuzzy input data for fuzzy inference to calculate the fuzzy output value of the estimated leakage amount; Step 35, by calculating the barycentric position on the fuzzy output membership curve, the specific estimated leakage amount after defuzzification is obtained. The specific process of step 5 is:

9. The adaptive valley hydrocarbon diffusion barrier method of claim 5, wherein, Step 51, collect data related to the device, filter and convert the data, then pre-process the data, and then divide the data set into a training set and a test set; ​ Step 52, selecting a long short-term memory network and a gated recurrent unit in the recurrent neural network architecture to capture long-term dependencies in the sequence data; Step 53, constructing an RNN model and defining the input layer, hidden layer, and output layer of the RNN; Step 54, inputting the training set into the RNN model, performing backpropagation and weight updating to adjust the model parameters, and obtaining a trained RNN model.

Images