Flooding-proof system of fire pump room

Abstract

The invention discloses an anti-flooding system for a fire pump room, and relates to the technical field of anti-flooding equipment, and the system comprises a data collection module which is configured to collect data information outside the pump room, data information inside the pump room and monitoring project threshold data information; the data processing module is electrically connected with the data acquisition module, and is configured to verify, clean, convert, calculate, analyze, store and manage the data information outside the pump room and the data information inside the pump room; and the abnormity early warning module is electrically connected with the data processing module, and is configured to compare and analyze the data information outside the pump room, the data information inside the pump room and the monitoring item threshold data information to obtain an early warning result. Real-time monitoring and early warning can be achieved, the emergency response speed is increased under the emergency condition, safety of fire-fighting facilities and equipment in a fire-fighting pump room is guaranteed, data information outside the pump room and data information inside the pump room can be monitored in real time, automatic control can be conducted in time after monitoring and early warning, and normal operation of the fire-fighting pump room equipment is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of flood prevention equipment, in particular to a fire pump room flood prevention system. Background Art

[0002] With the acceleration of urbanization, high-rise buildings and large public buildings are increasing. Fire safety has become an important part of urban safety. The fire pump room is the core component of the building's fire-fighting equipment, providing a stable water source for the fire water system. When a fire occurs, the fire pump room should be able to start in time to provide sufficient water for fire fighting, effectively reducing casualties and property losses.

[0003] At present, the measures for preventing flooding in fire pump rooms are generally: pre-designing and marking the water level, and setting up steps of a certain height at the entrance of the pump room. However, these measures have the following disadvantages:

[0004] In daily operations, stairs increase the difficulty of entering and exiting, especially for workers carrying tools or equipment, increasing physical exertion and time costs. In particular, they increase the difficulty of carrying large equipment or tools, which may require additional manpower or equipment, increasing maintenance costs and time;

[0005] In an emergency, stairs may delay entry and exit, affecting emergency response speed;

[0006] The lack of real-time monitoring and early warning makes it impossible to detect water level anomalies in time, delaying emergency response. In response to this, we have proposed a fire pump room flood prevention system. Summary of the Invention

[0007] In order to solve the above technical problems, a fire pump room flood prevention system is provided. This technical solution solves the above problems.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: a fire pump room flood prevention system, comprising:

[0009] A data acquisition module is configured to collect data information outside the pump room, data information inside the pump room, and threshold data information of monitoring items;

[0010] The data processing module is electrically connected to the data acquisition module and is configured to verify, clean, convert, calculate, analyze, store and manage the data information outside the pump room and the data information inside the pump room;

[0011] The abnormal warning module is electrically connected to the data processing module and is configured to compare and analyze the data information outside the pump room, the data information inside the pump room and the monitoring item threshold data information to obtain a warning result;

[0012] The automatic control module is electrically connected to the abnormal warning module and is configured to automatically control the equipment in the pump room based on the warning result;

[0013] The remote monitoring module is electrically connected to the data acquisition module, the data processing module, the abnormal warning module and the automatic control module respectively, and is used to obtain the data information outside the pump room, the data information inside the pump room and the warning results, and remotely control the equipment in the pump room.

[0014] Preferably, the report generation module is electrically connected to the data acquisition module, the data analysis module, the abnormal warning module, the automatic control module and the remote monitoring module, and is used to record and store the data information outside the pump room, the data information inside the pump room, the warning results and the automatic control process data information, generate corresponding reports and statistical data, and view them in real time based on the remote monitoring module.

[0015] Preferably, the data information outside the pump room includes flooding information and image information; the data information inside the pump room includes flooding information and image information; wherein the data information outside the pump room is monitored by a water immersion alarm arranged in front of the pump room door; the data information inside the pump room is monitored by a water immersion alarm arranged inside the pump room; the image information in front of the pump room door is monitored by a high-definition hemispherical camera arranged in front of the pump room door; and both the inside and outside of the pump room are monitored by hemispherical high-definition cameras.

[0016] Preferably, the specific steps for verifying, cleaning, converting, calculating, analyzing, storing and managing data information are as follows:

[0017] Conduct format compliance check and implementation validity verification on data information outside the pump room and data information inside the pump room;

[0018] For outlier processing, a flood depth threshold is set to automatically filter out sudden changes that exceed the physical limit, and the mean filling method of adjacent time periods is used to handle missing values;

[0019] Convert the data information format to a consistent one, standardize the units, and unify the measurement units of different sensors;

[0020] Calculate flooding trends in real time, calculate the water level rise rate based on a differential algorithm, combine historical data to predict flooding conditions in the next 10 minutes, extract image feature values, calculate the proportion of water flow pixels and flow velocity vector parameters in the image, and quantify the flooding severity value;

[0021] Analyze the frequency of flooding outside the pump house during rainy season based on time series algorithm;

[0022] The processed data information is stored in real time based on the database.

[0023] Preferably, the steps for calculating the water level rise rate using the differential algorithm are:

[0024] Real-time water level data is obtained through the flood alarm outside the pump room, with the water level and timestamp recorded every minute. Historical water level data for the same period of the rainy season over the past three years is obtained and stored by date and time period to form a historical database.

[0025] Select the water level data of the last five consecutive time points, calculate the water level difference between adjacent time points, average the difference, and divide it by the time interval value to obtain the water level rise rate per unit time;

[0026] The steps to predict flooding in the next 10 minutes based on historical data are as follows:

[0027] Compare the currently calculated water level rise rate with the historical data for the same period, filter out the data for the historical time period with the smaller deviation from the current rate, assign different weights to the data according to the time distance, with the recent data having a higher weight. After weighted calculation, predict the water level growth value in the next 10 minutes, generate a prediction curve and mark the time point corresponding to the warning threshold.

[0028] Preferably, the steps for quantifying the flood severity value are:

[0029] Based on real-time video capture from a high-definition hemispherical camera, the color image is converted into a grayscale image, and Gaussian filtering is used to remove noise and correct image distortion.

[0030] The water flow pixel ratio is calculated by using a mixed Gaussian model to establish a ground background model. The current frame image is compared with the background model to extract the water flow foreground area. After binarization, the number of foreground pixels is counted and then divided by the total number of pixels in the area of ​​interest to obtain the water flow pixel ratio, which is used to reflect the flood coverage.

[0031] The velocity vector parameter calculation is based on the optical flow algorithm to calculate the pixel displacement vectors of adjacent video frames. After filtering the noise vector, the effective vectors are weighted averaged to obtain the average velocity vector. Based on the pixel physical resolution of the camera calibration, the pixel velocity is converted into the actual velocity.

[0032] The severity of flooding is quantified by setting weights for the proportion of water flow pixels, actual flow velocity and water level rise rate. After normalizing each indicator, the severity value of flooding is obtained by weighted calculation. The larger the value, the more serious the flooding situation.

[0033] Preferably, the specific warning steps in the abnormal warning module are:

[0034] Set different flooding depth thresholds and compare the real-time collected flooding depth outside the pump room with the preset different thresholds in real time. When the water level reaches the first threshold, a yellow warning is triggered, and when it reaches the second threshold, a red warning is triggered. Establish a historical baseline for flooding data inside the pump room, and trigger an abnormal warning when the real-time data deviates from the baseline by ±30%;

[0035] A decision tree algorithm is used to construct a flood risk assessment model. The input parameters include flood depth, rising rate and image water flow characteristics, and the output is risk level and disposal recommendations.

[0036] Preferably, the automatic control steps in the automatic control module are:

[0037] Based on different warning levels, hierarchical control is implemented. When a yellow warning is issued, the intelligent tracking function of the high-definition dome camera outside the pump room is automatically activated to continuously monitor the water flow. The control mode of the underground telescopic flood prevention threshold is switched to the ready-to-raise state.

[0038] When a red alert occurs, the underground telescopic flood prevention threshold will rise within 10 seconds. When the height reaches the preset flood prevention height, it will block external water from entering the pump room.

[0039] After the automatic control module sends the control command, it receives the equipment status feedback signal in real time. If no feedback is received within the specified time or the feedback is abnormal, it triggers the backup equipment to start and sends an equipment failure alarm to the on-duty personnel.

[0040] Preferably, the remote monitoring module is remotely controlled by carrying control software and supports B / S architecture access. The on-duty personnel of the fire control center log in to the system through the browser by entering the IP address for real-time viewing and remote control.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] The present invention proposes a fire pump room flood prevention system comprising: a flood alarm, a high-definition dome camera, an underground telescopic flood prevention threshold, a submersible sewage pump, a water inlet pipe solenoid valve, and a terminal computer. The terminal computer is connected to the flood alarm, high-definition dome camera, submersible sewage pump, and water inlet pipe solenoid valve, respectively, and is used to control the operating status of the submersible sewage pump, solenoid valve, and underground telescopic flood prevention threshold based on flooding information and images of the pump room interior and in front of the door. This system facilitates normal access for fire pump room personnel and the transport of equipment and tools, and enables real-time monitoring and early warning. This system improves emergency response speed in emergencies, ensures the safety of firefighting facilities and equipment within the fire pump room, and monitors data both outside and inside the pump room in real time. Following monitoring and early warning, automated control can be promptly implemented to ensure the normal operation of fire pump room equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A schematic diagram of the overall scheme of a fire pump room flood prevention system according to an embodiment of the present invention;

[0044] Figure 2 The figure is a flow chart of a fire pump room flood prevention system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0045] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are merely examples, and those skilled in the art may conceive of other obvious variations.

[0046] Reference Figure 1 and Figure 2 As shown, the fire pump room flood prevention system includes:

[0047] A data acquisition module is configured to collect data information outside the pump room, data information inside the pump room, and threshold data information of monitoring items;

[0048] The data processing module is electrically connected to the data acquisition module and is configured to verify, clean, convert, calculate, analyze, store and manage the data information outside the pump room and the data information inside the pump room;

[0049] The abnormal warning module is electrically connected to the data processing module and is configured to compare and analyze the data information outside the pump room, the data information inside the pump room and the monitoring item threshold data information to obtain a warning result;

[0050] The automatic control module is electrically connected to the abnormal warning module and is configured to automatically control the equipment in the pump room based on the warning result;

[0051] The remote monitoring module is electrically connected to the data acquisition module, the data processing module, the abnormal warning module and the automatic control module respectively, and is used to obtain the data information outside the pump room, the data information inside the pump room and the warning results, and remotely control the equipment in the pump room.

[0052] This application simultaneously collects data outside the pump room (flooding information, image information), inside the pump room (flooding information, image information), and threshold data to form a three-dimensional monitoring network. The flood alarm captures water level changes in real time, and high-definition cameras record on-site images. Threshold data (such as warning water levels and equipment operating parameters) provides a benchmark for subsequent analysis, avoiding monitoring blind spots caused by a single data source.

[0053] Threshold data for monitoring projects can be dynamically adjusted based on historical data or on-site conditions (e.g., raising flood warning thresholds during the rainy season), ensuring targeted and adaptable data collection and reducing false positives or missed reports.

[0054] Ensure data accuracy and reliability through verification (format, validity), cleaning (denoising, completion), and conversion (standardization). For example, we can eliminate abnormal data from flood alarms due to voltage fluctuations and standardize the units of different sensors to avoid decision-making errors caused by data confusion.

[0055] By using differential algorithms to calculate the rate of water level rise and extracting image features to quantify flood severity, the raw data is converted into information that can guide decision-making. Combined with historical data, the flood trend over the next 10 minutes is predicted, allowing time for emergency response.

[0056] Categorize and store real-time and historical data to form a pump room flood prevention database, providing data support for subsequent equipment maintenance and strategy optimization (such as analyzing the frequency of flooding during the rainy season and deploying flood prevention measures in advance);

[0057] Multi-dimensional comparison and warning: Simultaneously analyzes real-time data (flood depth, image water flow characteristics) and threshold data, triggering warnings through logical comparison (such as when the water level exceeds the threshold and the image recognizes water flow) to avoid false alarms caused by a single condition. When the flood alarm sounds but the camera does not capture water flow, the system automatically verifies the data to reduce false warnings caused by environmental interference.

[0058] The report generation module is electrically connected to the data acquisition module, data analysis module, abnormal warning module, automatic control module and remote monitoring module, and is used to record and store the data information outside the pump room, the data information inside the pump room, the warning results and the automatic control process data information, generate corresponding reports and statistical data, and view them in real time based on the remote monitoring module.

[0059] The fire pump room flood prevention system of the present application monitors the data information outside the pump room and the data information inside the pump room in real time, and performs early warning and automatic control according to the monitoring results, so as to discover and solve problems in advance, ensure the convenience of daily personnel passage and equipment maintenance, and the normal operation of the equipment, thereby ensuring the normal operation of the fire pump room and the fire extinguishing effect.

[0060] The data information outside the pump room includes flooding information and image information; the data information inside the pump room includes flooding information and image information; the data information outside the pump room is monitored by a water immersion alarm installed in front of the pump room door; the data information inside the pump room is monitored by a water immersion alarm installed inside the pump room; the image information in front of the pump room door is monitored by a high-definition hemispherical camera installed in front of the pump room door; both the inside and outside of the pump room are monitored by hemispherical high-definition cameras.

[0061] This application realizes multi-dimensional monitoring of flooding information and image information by deploying water flood alarms and high-definition hemispheric cameras outside and inside the pump room respectively, which not only improves the accuracy and comprehensiveness of monitoring, but also provides a sufficient basis for emergency response and daily management. It fundamentally improves the reliability and intelligence level of the fire pump room flood prevention system, ensures that the fire pump room can respond in a timely and effective manner when facing flooding risks, and ensures the normal operation of fire protection facilities.

[0062] The specific steps for verifying, cleaning, converting, calculating, analyzing, storing and managing data information are as follows:

[0063] Conduct format compliance check and implementation validity verification on data information outside the pump room and data information inside the pump room;

[0064] For outlier processing, a flood depth threshold is set to automatically filter out sudden changes that exceed the physical limit, and the mean filling method of adjacent time periods is used to handle missing values;

[0065] Convert the data information format to a consistent one, standardize the units, and unify the measurement units of different sensors;

[0066] Calculate flooding trends in real time, calculate the water level rise rate based on a differential algorithm, combine historical data to predict flooding conditions in the next 10 minutes, extract image feature values, calculate the proportion of water flow pixels and flow velocity vector parameters in the image, and quantify the flooding severity value;

[0067] Analyze the frequency of flooding outside the pump house during rainy season based on time series algorithm;

[0068] The processed data information is stored in real time based on the database.

[0069] This application can improve data quality by verifying format compliance and validity, unifying data standards, filtering invalid data, and improving integrity through outlier and missing value processing. Secondly, it standardizes data formats, eliminates unit confusion barriers, and enhances usability to support cross-system integration. At the same time, it can calculate and analyze flooding trends in real time, combine differential algorithms and historical data to predict the situation in the next 10 minutes, quantify image features to obtain severity values, and analyze the frequency of flooding in the rainy season.

[0070] The steps for calculating the water level rise rate using the differential algorithm are:

[0071] Real-time water level data is obtained through the flood alarm outside the pump room, with the water level and timestamp recorded every minute. Historical water level data for the same period of the rainy season over the past three years is obtained and stored by date and time period to form a historical database.

[0072] Select the water level data of the last five consecutive time points, calculate the water level difference between adjacent time points, average the difference, and divide it by the time interval value to obtain the water level rise rate per unit time;

[0073] The steps to predict flooding in the next 10 minutes based on historical data are as follows:

[0074] Compare the currently calculated water level rise rate with the historical data for the same period, filter out the data for the historical time period with the smaller deviation from the current rate, assign different weights to the data according to the time distance, with the recent data having a higher weight. After weighted calculation, predict the water level growth value in the next 10 minutes, generate a prediction curve and mark the time point corresponding to the warning threshold.

[0075] The differential algorithm used in this application calculates the rate of water level rise and combines it with historical data to predict future flooding conditions. It has significant advantages in pump room water level monitoring and early warning. The differential algorithm collects water level data and timestamps every minute, takes the water level data of the last five consecutive time points, calculates the average of adjacent differences and divides it by the time interval, and obtains the water level rise rate in real time. It can not only respond quickly to water level changes, but also filter out abnormal fluctuations. The calculation process is simple and adapts to real-time monitoring needs. The prediction method based on historical data compares the current water level rise rate with historical data of the same period, selects historical period data with small deviations, and assigns weights according to the distance in time. Recent data has a higher weight, and the weighted calculation predicts the water level growth value in the next 10 minutes to generate a prediction curve with the warning threshold marked. This method refers to the trends under similar hydrological conditions, integrates long-term laws and recent changes, and makes predictions more accurate.

[0076] The steps to quantify the flood severity value are:

[0077] Based on real-time video capture from a high-definition hemispherical camera, the color image is converted into a grayscale image, and Gaussian filtering is used to remove noise and correct image distortion.

[0078] The water flow pixel ratio is calculated by using a mixed Gaussian model to establish a ground background model. The current frame image is compared with the background model to extract the water flow foreground area. After binarization, the number of foreground pixels is counted and then divided by the total number of pixels in the area of ​​interest to obtain the water flow pixel ratio, which is used to reflect the flood coverage.

[0079] The velocity vector parameter calculation is based on the optical flow algorithm to calculate the pixel displacement vectors of adjacent video frames. After filtering the noise vector, the effective vectors are weighted averaged to obtain the average velocity vector. Based on the pixel physical resolution of the camera calibration, the pixel velocity is converted into the actual velocity.

[0080] The severity of flooding is quantified by setting weights for the proportion of water flow pixels, actual flow velocity and water level rise rate. After normalizing each indicator, the severity value of flooding is obtained by weighted calculation. The larger the value, the more serious the flooding situation.

[0081] In the image preprocessing stage, this application converts the color image into a grayscale image, removes noise and corrects distortion through Gaussian filtering to ensure the quality of subsequent analysis data. The water flow pixel ratio calculation uses a mixed Gaussian model to extract the water flow foreground, and the ratio is obtained through binarization statistics to intuitively present the flood coverage and change trend. The flow velocity vector parameter calculation is based on the optical flow algorithm. After filtering the noise, it is combined with camera calibration to convert it into the actual flow velocity, reflecting the flushing force and direction of the water flow.

[0082] The specific warning steps in the abnormal warning module are:

[0083] Set different flooding depth thresholds and compare the real-time collected flooding depth outside the pump room with the preset different thresholds in real time. When the water level reaches the first threshold, a yellow warning is triggered, and when it reaches the second threshold, a red warning is triggered. Establish a historical baseline for flooding data inside the pump room, and trigger an abnormal warning when the real-time data deviates from the baseline by ±30%;

[0084] A decision tree algorithm is used to construct a flood risk assessment model. The input parameters include flood depth, rising rate and image water flow characteristics, and the output is risk level and disposal recommendations.

[0085] The series of steps of the abnormal warning module of this application provide a multi-level and precise solution for the pump room flooding risk warning through threshold comparison, baseline monitoring and intelligent algorithm evaluation. Different flooding depth thresholds are set for real-time comparison. The graded warning mechanism (yellow and red warnings) can intuitively and clearly reflect the degree of risk, making it easier for management personnel to respond according to the severity, such as strengthening inspections during yellow warnings and immediately launching emergency plans during red warnings.

[0086] The automatic control steps in the automatic control module are:

[0087] Based on different warning levels, hierarchical control is implemented. When a yellow warning is issued, the intelligent tracking function of the high-definition dome camera outside the pump room is automatically activated to continuously monitor the water flow. The control mode of the underground telescopic flood prevention threshold is switched to the ready-to-raise state.

[0088] When a red alert occurs, the underground telescopic flood prevention threshold will rise within 10 seconds. When the height reaches the preset flood prevention height, it will block external water from entering the pump room.

[0089] After the automatic control module sends the control command, it receives the equipment status feedback signal in real time. If no feedback is received within the specified time or the feedback is abnormal, it triggers the backup equipment to start and sends an equipment failure alarm to the on-duty personnel.

[0090] The remote monitoring module is remotely controlled by the equipped control software and supports B / S architecture access. The on-duty personnel of the fire control center log in to the system through the browser by entering the IP address for real-time viewing and remote control.

[0091] The specific control methods are:

[0092] Specifically, under normal circumstances, the underground telescopic anti-flooding threshold is retracted into the ground, with the top flush with the ground, which is convenient for personnel and equipment to enter and exit; when there is water intrusion alarm outside the pump room, the abnormal warning module will issue an abnormal flooding warning, and the high-definition hemispheric camera outside the pump room will shoot the front of the door. When the image recognizes that there is water flow on the ground outside the pump room, the underground telescopic anti-flooding threshold will rise to block the water from flowing into the fire pump room, and the special submersible sewage pump outside the pump room will be turned on. When the image recognizes that there is no water flow, the status quo is maintained. After the on-duty personnel of the fire control room go to the scene to confirm that the flooding is effectively controlled, the alarm is lifted, and the underground telescopic anti-flooding threshold and the special submersible sewage pump outside the pump room are turned on. The water sewage pump returns to normal working condition; when there is water intrusion alarm in the pump room, the abnormal warning module issues a flooding abnormal warning, and the high-definition hemispheric camera in the pump room shoots the interior. When the image recognizes that water in the pump room overflows the drainage ditch, the underground telescopic anti-flood threshold rises to block the water from flowing out of the fire pump room, the solenoid valve of the water inlet pipe is closed, and all the submersible sewage pumps in the pump room are turned on. When the image recognizes that there is no water flow on the ground in the pump room, the status quo is maintained. After the on-duty personnel of the fire control room go to the scene to confirm that the flooding is effectively controlled, the alarm is lifted, and the underground telescopic anti-flood threshold and the submersible sewage pump in the pump room return to normal working condition.

[0093] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. Various changes and improvements are possible without departing from the spirit and scope of the present invention, and such changes and improvements fall within the scope of the invention as claimed.

Claims

1. Fire pump room flood prevention system, characterized by: include: A data acquisition module is configured to collect data information outside the pump room, data information inside the pump room, and threshold data information of monitoring items; The data processing module is electrically connected to the data acquisition module and is configured to verify, clean, convert, calculate, analyze, store and manage the data information outside the pump room and the data information inside the pump room; The abnormal warning module is electrically connected to the data processing module and is configured to compare and analyze the data information outside the pump room, the data information inside the pump room and the monitoring item threshold data information to obtain a warning result; The automatic control module is electrically connected to the abnormal warning module and is configured to automatically control the equipment in the pump room based on the warning result; The remote monitoring module is electrically connected to the data acquisition module, the data processing module, the abnormal warning module and the automatic control module respectively, and is used to obtain the data information outside the pump room, the data information inside the pump room and the warning results, and remotely control the equipment in the pump room.

2. The fire pump room flood prevention system according to claim 1, characterized in that: It also includes a report generation module, which is electrically connected to the data acquisition module, data analysis module, abnormal warning module, automatic control module and remote monitoring module, and is used to record and store the data information outside the pump room, data information inside the pump room, warning results and automatic control process data information, generate corresponding reports and statistical data, and view them in real time based on the remote monitoring module.

3. The fire pump room flood prevention system according to claim 1, characterized in that: The data information outside the pump room includes flooding information and image information; the data information inside the pump room includes flooding information and image information; the data information outside the pump room is monitored by a water immersion alarm installed in front of the pump room door; the data information inside the pump room is monitored by a water immersion alarm installed inside the pump room; the image information in front of the pump room door is monitored by a high-definition hemispherical camera installed in front of the pump room door; both the inside and outside of the pump room are monitored by hemispherical high-definition cameras.

4. The fire pump room flood prevention system according to claim 1, characterized in that: The specific steps for verifying, cleaning, converting, calculating, analyzing, storing and managing data information are as follows: Conduct format compliance check and implementation validity verification on data information outside the pump room and data information inside the pump room; For outlier processing, a flood depth threshold is set to automatically filter out sudden changes that exceed the physical limit, and the mean filling method of adjacent time periods is used to handle missing values; Convert the data information format to a consistent one, standardize the units, and unify the measurement units of different sensors; Calculate flooding trends in real time, calculate the water level rise rate based on a differential algorithm, combine historical data to predict flooding conditions in the next 10 minutes, extract image feature values, calculate the proportion of water flow pixels and flow velocity vector parameters in the image, and quantify the flooding severity value; Analyze the frequency of flooding outside the pump house during rainy season based on time series algorithm; The processed data information is stored in real time based on the database.

5. The fire pump room flood prevention system according to claim 4, characterized in that: The steps for calculating the water level rise rate using the differential algorithm are: Real-time water level data is obtained through the flood alarm outside the pump room, with the water level and timestamp recorded every minute. Historical water level data for the same period of the rainy season over the past three years is obtained and stored by date and time period to form a historical database. Select the water level data of the last five consecutive time points, calculate the water level difference between adjacent time points, average the difference, and divide it by the time interval value to obtain the water level rise rate per unit time; The steps to predict flooding in the next 10 minutes based on historical data are as follows: Compare the currently calculated water level rise rate with the historical data for the same period, filter out the data for the historical time period with the smaller deviation from the current rate, assign different weights to the data according to the time distance, with the recent data having a higher weight. After weighted calculation, predict the water level growth value in the next 10 minutes, generate a prediction curve and mark the time point corresponding to the warning threshold.

6. The fire pump room flood prevention system according to claim 4, characterized in that: The steps to quantify the flood severity value are: Based on real-time video capture from a high-definition hemispherical camera, the color image is converted into a grayscale image, and Gaussian filtering is used to remove noise and correct image distortion. The water flow pixel ratio is calculated by using a mixed Gaussian model to establish a ground background model. The current frame image is compared with the background model to extract the water flow foreground area. After binarization, the number of foreground pixels is counted and then divided by the total number of pixels in the area of ​​interest to obtain the water flow pixel ratio, which is used to reflect the flood coverage. The velocity vector parameter calculation is based on the optical flow algorithm to calculate the pixel displacement vectors of adjacent video frames. After filtering the noise vector, the effective vectors are weighted averaged to obtain the average velocity vector. Based on the pixel physical resolution of the camera calibration, the pixel velocity is converted into the actual velocity. The severity of flooding is quantified by setting weights for the proportion of water flow pixels, actual flow velocity and water level rise rate. After normalizing each indicator, the severity value of flooding is obtained by weighted calculation. The larger the value, the more serious the flooding situation.

7. The fire pump room flood prevention system according to claim 1, characterized in that: The specific warning steps in the abnormal warning module are: Set different flooding depth thresholds and compare the real-time collected flooding depth outside the pump room with the preset different thresholds in real time. When the water level reaches the first threshold, a yellow warning is triggered, and when it reaches the second threshold, a red warning is triggered. Establish a historical baseline for flooding data inside the pump room, and trigger an abnormal warning when the real-time data deviates from the baseline by ±30%; A decision tree algorithm is used to construct a flood risk assessment model. The input parameters include flood depth, rising rate and image water flow characteristics, and the output is risk level and disposal recommendations.

8. The fire pump room flood prevention system according to claim 1, characterized in that: The automatic control steps in the automatic control module are: Based on different warning levels, hierarchical control is implemented. When a yellow warning is issued, the intelligent tracking function of the high-definition dome camera outside the pump room is automatically activated to continuously monitor the water flow. The control mode of the underground telescopic flood prevention threshold is switched to the ready-to-raise state. When a red alert occurs, the underground telescopic flood prevention threshold will rise within 10 seconds. When the height reaches the preset flood prevention height, it will block external water from entering the pump room. After the automatic control module sends the control command, it receives the equipment status feedback signal in real time. If no feedback is received within the specified time or the feedback is abnormal, it triggers the backup equipment to start and sends an equipment failure alarm to the on-duty personnel.

9. The fire pump room flood prevention system according to claim 1, characterized in that: The remote monitoring module is remotely controlled by the equipped control software and supports B / S architecture access. The on-duty personnel of the fire control center log in to the system through the browser by entering the IP address for real-time viewing and remote control.

Images