Manhole mud level detection system and method based on multimodal perception fusion

Abstract

The present application provides a manhole mud level detection system and method based on multimodal sensing fusion. The system comprises: a basic data acquisition module for collecting basic well condition data; a mud level measurement module for measuring manhole mud level data based on the basic well condition data. The manhole mud level data includes bottomhole mud level data for measuring bottomhole silt and mud level data at multiple well nodes within the well floc layer measured using a multimodal sensor; a data processing module for fusing the mud level data from multiple well nodes measured by the multimodal sensor to obtain fused mud level data for the well floc layer. The bottomhole mud level data and the fused mud level data constitute the manhole mud level data. The present application can automatically identify water level, mud level, and well height, adapt to variable hydraulic conditions, and detect mud levels at multiple key locations within the manhole. It is portable, highly accurate, interference-resistant, and reusable.

Description

Technical Field

[0001] The present application belongs to the technical field of municipal drainage network monitoring, and relates to a municipal drainage network system, and in particular to a manhole mud level detection system and method based on multimodal perception fusion. Background Art

[0002] In municipal drainage systems, sludge deposits within manholes and drainage pipes have become a core issue hindering network performance. The continuous accumulation of sludge not only directly reduces the flow-through cross-section (for example, in a 1200mm diameter pipe, the flow capacity decreases by 42% when 30cm of sludge accumulates), but also triggers complex hydraulic hazards during the rainy season, including flow distortion, vortex surges, and overflow pollution. Furthermore, an analysis of the material composition of deposited sludge shows that volatile solids generally account for over 60%. Anaerobic hydrolysis continuously releases toxic gases such as H2S (at concentrations of 200-800ppm), leading to corrosion rates 3-5 times higher than under normal operating conditions, significantly shortening the service life of the pipeline network. These systemic risks urgently require effective management through precise mud level monitoring technology.

[0003] Although the current "Technical Regulations for Operation, Maintenance and Safety of Urban Drainage Pipes and Pumping Stations" (CJJ68-2016) stipulates a limit on the depth of mud accumulation in inspection wells (≤50mm for well chambers with sedimentation tanks, ≤1 / 5 of the pipe diameter for well chambers without sedimentation tanks), existing measurement technologies have significant limitations: manual detection methods have a low coverage rate (single-point measurement accounts for more than 80%) and pose personnel safety risks; the hanging hammer method is limited by the irregular structure of the well chamber (irregular wells account for more than 35%), and the measurement error often exceeds ±15%; static ultrasonic equipment is easily affected by turbulent interference in dynamic flow fields (the error rate increases by 2.3 times when the flow rate is greater than 0.3m / s). Actual measured data show that the data drift amplitude can reach 8-12cm under turbulent conditions. In addition, traditional equipment has fundamental defects in identifying the transition layer at the mud-water interface. The floc transition layer with a thickness of 5-20 cm has multiphase flow characteristics (density gradient 1.02-1.15 g / cm³), which causes the ultrasonic echo amplitude to attenuate by 40-60 dB and the laser scattering angle deviation to reach 15°-25°. This is the core technical bottleneck of the current mud level measurement with an average error of ±5 cm.

[0004] The technical challenges are specifically manifested in the following three aspects:

[0005] First, the variation of optical / acoustic properties of the floc transition layer leads to interface blurring effect;

[0006] Second, the unstructured well chamber environment results in insufficient spatial adaptability of mechanical measurement devices;

[0007] Third, multiphase flow interference under dynamic hydraulic conditions causes signal distortion.

[0008] For example, equipment using multi-frequency ultrasonic composite detection still exhibits a 12-18mm deviation in media with an oil content greater than 3%. These deficiencies result in a lack of reliable data support for quantitative assessment of the siltation status of drainage networks, severely hindering the modification of hydraulic models and the optimization of operation and maintenance decisions.

[0009] Therefore, there is an urgent need for a mud level measurement technology that has dynamic multi-point detection capabilities, strong anti-interference ability, and can accurately identify the floc transition layer. The measurement accuracy can be improved through multi-physical field fusion perception and intelligent algorithms to provide highly reliable mud level benchmark data for drainage network operation. Summary of the Invention

[0010] The purpose of this application is to provide a manhole mud level detection system and method based on multimodal perception fusion, which is used to solve the problem of large error in mud level measurement of manholes in the prior art.

[0011] In the first aspect, the present application provides a mud level detection system for an inspection well based on multimodal sensing fusion, the system comprising: a basic data acquisition module for collecting basic well condition data; a mud level measurement module for measuring the mud level data of the inspection well based on the basic well condition data, the mud level data of the inspection well comprising bottom hole mud level data for measuring the bottom hole deposited silt and mud level data of multiple well nodes of the floc layer in the well measured based on a multimodal sensor; a data processing module for fusing the mud level data of multiple well nodes measured by the multimodal sensor to obtain mud level fusion data of the floc layer in the well; the bottom hole mud level data and the mud level fusion data constitute the inspection well mud level data.

[0012] In an implementation of the first aspect, the basic well condition data includes well height data and water level data; the basic data acquisition module includes: a well height measurement unit, a water level measurement unit and a measurement preparation unit; the well height measurement unit includes an electric well height detection telescopic rod, which obtains the well height data of the inspection well by controlling the well height detection telescopic rod to move downward to contact the bottom of the well; the water level measurement unit includes a water level detection sensor installed at the wellhead of the inspection well, which is used to measure the vertical distance from the wellhead to the water surface, and obtain water level data based on the vertical distance and the well height data; the measurement preparation unit is used to control the measurement position of the mud level measurement module according to the water level data.

[0013] In one implementation of the first aspect, the mud level measurement module includes: a sedimentation silt measurement unit, a floc layer measurement unit, and a result output unit. The sedimentation silt measurement unit includes a telescopic mud level detection rod and a variable frequency ultrasonic detector. The telescopic mud level detection rod drives the variable frequency ultrasonic detector up and down by extending and retracting. The variable frequency ultrasonic detector detects the position of sedimentation silt at the bottom of the well using low-frequency ultrasound to obtain bottom-hole mud level data. The floc layer measurement unit includes a rotary motor, a transverse telescopic rod, and a laser detector mounted at the bottom end of the transverse telescopic rod. The variable frequency ultrasonic detector and the laser detector form the multimodal sensor. The transverse telescopic rod is connected to the telescopic mud level detection rod at a first end, and the variable frequency ultrasonic detector and the laser detector are mounted at a second end. The rotary motor drives the horizontal extension and retraction of the transverse telescopic rod. The high-frequency ultrasound detection of the variable frequency ultrasonic detector and the laser detector are used to measure mud level data at multiple well nodes in the floc layer within the well. The result output unit is configured to transmit the acquired bottom-hole mud level data and mud level data at multiple well nodes in the floc layer to the data processing module.

[0014] In an implementation of the first aspect, the basic data acquisition module, the mud level measurement module and the data processing module are installed in a box; the box is arranged above the inspection well, and the basic data acquisition module and the mud level measurement module can be folded and retracted in the box, and can be extended from the bottom of the box to perform data acquisition and measurement; the variable frequency ultrasonic detector and the laser detector are provided with a protective component on the outside; the protective component includes: a shock absorber unit, an anti-turbulence deflector and a self-cleaning brush; wherein, the shock absorber unit and the self-cleaning brush are arranged at the second end of the transverse telescopic rod; the shock absorber unit and the self-cleaning brush are arranged in the space formed by the anti-turbulence deflector.

[0015] In an implementation of the first aspect, the data processing module includes: a data set construction unit, a data processing unit, and a data storage unit; the data set construction unit constructs a data set based on the mud level data of multiple well nodes in the floc layer; the data processing unit performs fusion processing on the mud level data of multiple well nodes measured by the multimodal sensor in the data set by smoothing the data set and configuring weight coefficients for the mud level data collected by the variable frequency ultrasonic detector and the laser detector in the data set.

[0016] In an implementation of the first aspect, smoothing the data set includes smoothing the continuously collected mud level data of the floc layer using an exponentially weighted moving average calculation method; wherein the calculation formula of the exponentially weighted moving average calculation method is:

[0017] ,

[0018] in, represents the smoothed mud level data value at time t; express The data value of mud level containing flocculent layer at any moment; Indicates time The smoothed mud level data value; Represents the smoothing coefficient, with a value range of (0, 1), which is used to control the weight of new and old data.

[0019] In an implementation of the first aspect, configuring a weight coefficient for the mud level data collected by the variable-frequency ultrasonic detector and the laser detector in the data set includes: respectively calculating the coefficient of variation of the mud level data collected by the variable-frequency ultrasonic detector and the laser detector; and configuring a weight coefficient for the mud level data collected by the variable-frequency ultrasonic detector and the laser detector based on each of the coefficients of variation and a preset variation threshold.

[0020] In an implementation of the first aspect, the calculation formula of the coefficient of variation is:

[0021] ,

[0022] ,

[0023] ,

[0024] in, Represents the standard deviation of the data set; represents the mean of the data set; Indicates the measurements; Indicates the number of measurement samples; when the coefficient of variation CV value is less than the variation threshold, a preset high numerical weight is configured for the mud level data; when the coefficient of variation CV value is greater than the variation threshold, a preset low numerical weight is configured for the mud level data.

[0025] In an implementation of the first aspect, a calculation formula for configuring the weight of the mud level data of the variable ultrasonic detector is:

[0026] ,

[0027] The calculation formula for the mud level data weight of the laser detector is:

[0028] ,

[0029] The calculation formula for fusing the mud level data of multiple well nodes measured by the multimodal sensor in the data set is:

[0030] ,

[0031] in, Indicates the mud level fusion data of the floc layer; Represents the weight value of the mud level data of the variable frequency ultrasonic detector; Indicates the weight value of the mud level data of the laser detector; Represents the smoothed value of the mud level data of the variable frequency ultrasonic detector; Indicates the smoothed value of the mud level data of the laser detector; Indicates the coefficient of variation of the mud level data of the variable frequency ultrasonic detector; Represents the coefficient of variation of the mud level data of the laser detector.

[0032] In the second aspect, the present application provides a method for detecting mud level in an inspection well based on multimodal sensing fusion, the method comprising: collecting basic data on well conditions; measuring mud level data of the inspection well based on the basic data on well conditions, the measured mud level data of the inspection well comprising bottom hole mud level data of the bottom hole deposited silt and mud level data of multiple well position nodes of the floc layer in the well measured based on a multimodal sensor; fusing the mud level data of the multiple well position nodes measured by the multimodal sensor to obtain mud level fusion data of the floc layer in the well; the bottom hole mud level data and the mud level fusion data constitute the inspection well mud level data.

[0033] As described above, the inspection well mud level detection system and control method based on multimodal perception fusion described in this application have the following beneficial effects:

[0034] (1) The inspection well mud level detection system based on multimodal sensing fusion provided in this application can effectively reduce the interference of disturbances, bubbles, impurities, etc. in the well on the detection signal through multi-source detection and dynamic weighting algorithm, thereby improving the stability and accuracy of mud level and floc layer thickness measurement; it can automatically identify water level, mud level and well height, adapt to variable hydraulic conditions, and realize mud level detection at multiple key positions in the inspection well, with the advantages of portability, high precision, anti-interference and reusability; at the same time, it also greatly reduces the complexity of operation and labor intensity of personnel;

[0035] (2) The inspection well mud level detection method based on multimodal perception fusion provided in this application can intelligently adjust the detection mode according to different inspection well sizes, water levels, and liquid characteristics, and has good compatibility and strong adaptability; it can also accurately measure the sludge height of the drainage pipe, reduce manual judgment errors, and improve the maintenance efficiency of the drainage system. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1Shown is a schematic diagram of the overall structural framework of the inspection well mud level detection system based on multimodal perception fusion described in an embodiment of the present application.

[0037] Figure 2 Shown is a structural diagram of the inspection well mud level detection system based on multimodal perception fusion described in an embodiment of the present application.

[0038] Figure 3 Shown is a schematic diagram of the main structure of the inspection well mud level detection system based on multimodal perception fusion described in an embodiment of the present application.

[0039] Figure 4 Shown is a schematic diagram of on-site equipment deployment of an inspection well mud level detection device based on multimodal perception fusion in one embodiment of the present application.

[0040] Figure 5 Shown is a structural schematic diagram of the box and control components of an inspection well mud level detection device based on multimodal perception fusion described in an embodiment of the present application.

[0041] Figure 6 Shown is a structural diagram of the intelligent monitoring and rotation components of an inspection well mud level detection device based on multimodal perception fusion as described in an embodiment of the present application.

[0042] Figure 7 Shown is a structural schematic diagram of the protection components in one embodiment of the inspection well mud level detection method based on multimodal perception fusion described in an embodiment of the present application.

[0043] Figure 8 Shown is a flow chart of an embodiment of the method for detecting mud level in an inspection well based on multimodal perception fusion as described in an embodiment of the present application.

[0044] Figure 9 Shown is a schematic diagram of the operational flow of an inspection well mud level detection method based on multimodal perception fusion in one embodiment of the present application.

[0045] Component number description

[0046] 1 Manhole mud level detection system based on multimodal perception fusion 11 Basic data acquisition module 111 Well height measurement unit 112 Water level measurement unit 113 Measurement preparation unit 12 Mud level measurement module 121 Sedimentation and silt measurement unit 122 Floc layer measurement unit 123 Result output unit 13 Data processing module 131 Dataset building block 132 Data processing unit 133 Data storage unit 2 Cabinet 3 Control components 31 base 32 power supply 33 Electric telescopic rod set 34 Water level ultrasonic survey probe 4 Intelligent monitoring and rotating components 41 Well depth detection telescopic rod 42 Mud level detection telescopic rod 43 Rotating electric machines 44 Horizontal telescopic rod 45 Variable frequency ultrasonic detector 46 Laser detectors 5 Protective components 51 shock absorber unit 52 Anti-turbulence deflector 53 Self-cleaning brush 6 server DETAILED DESCRIPTION

[0047] The following describes the embodiments of the present application through specific examples. Those skilled in the art can easily understand the other advantages and effects of the present application from the content disclosed in this specification. The present application can also be implemented or applied through other different specific embodiments. The details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments can be combined with each other unless they conflict.

[0048] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application. Therefore, the illustrations only show components related to the present application and are not drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component can be changed at will, and the component layout type may also be more complicated.

[0049] The inspection well mud level detection system based on multimodal perception fusion provided in this application can automatically identify water level, mud level and well height, adapt to variable hydraulic conditions, and realize mud level detection at multiple key positions in the inspection well. It has the advantages of portability, high precision, anti-interference and reusability.

[0050] The following embodiments of the present application provide an inspection well mud level detection system based on multimodal sensing fusion, which solves the problems of insufficient spatial adaptability of existing mud level measurement technology in complex well chamber environments, signal distortion under dynamic hydraulic conditions, and fuzzy effect of the floc transition layer interface.

[0051] like Figures 1 to 3 As shown, the overall structural framework diagram of the inspection well mud level detection system based on multimodal perception fusion described in the embodiment of the present application specifically includes: a basic data acquisition module 11, a mud level measurement module 12, and a data processing module 13. The inspection well mud level detection system 1 based on multimodal perception fusion provided in the following embodiment of the present application can automatically identify the water level, mud level and well height through the coordinated work of these three modules, adapt to variable hydraulic conditions, and realize mud level detection at multiple key positions in the inspection well. It has the characteristics of portability, high precision, anti-interference and reusability.

[0052] See also Figures 4 to 7, respectively showing a schematic diagram of on-site equipment deployment of the inspection well mud level detection device based on multimodal perception fusion in one embodiment of the present application, a structural schematic diagram of the box and control components of the inspection well mud level detection device based on multimodal perception fusion in one embodiment of the present application, a structural schematic diagram of the intelligent monitoring and rotation components of the inspection well mud level detection device based on multimodal perception fusion in one embodiment of the present application, and a structural schematic diagram of the protection component of the inspection well mud level detection method based on multimodal perception fusion in one embodiment of the present application.

[0053] In this embodiment, the basic data acquisition module, the mud level measurement module and the data processing module are installed in a box 2; the box 2 is set above the inspection well, and the basic data acquisition module 11 and the mud level measurement module 12 can be folded and retracted in the box 2, and can be extended from the bottom of the box 2 for data acquisition and measurement.

[0054] Specifically, the box body 2 is further provided with a control component 3, an intelligent monitoring and rotation component 4, and a protection component 5.

[0055] The control component 3 is suspended and rotated above the inspection well through a foldable base 31, and the control component 3 includes a power supply 32, a multi-stage nested electric telescopic rod group 33, and a water level ultrasonic survey probe 34.

[0056] The intelligent monitoring and rotation component 4 is disposed in the box 2 and connected to the control component 3 .

[0057] In this embodiment, the intelligent monitoring and rotation assembly 4 includes: a well depth detection telescopic rod 41, a mud level detection telescopic rod 42, a rotary motor 43, a transverse telescopic rod 44, a variable frequency ultrasonic detector 45, and a laser detector 46. The well depth detection telescopic rod 41 and the mud level detection telescopic rod 42 are disposed within the housing 2; a rotary motor 43 is disposed at the end of the mud level detection telescopic rod 42, which controls the horizontal rotation of the transverse telescopic rod 44; a variable frequency ultrasonic detector 45 and a laser detector 46 are disposed at the end of the transverse telescopic rod 44; a first end of the transverse telescopic rod 44 is connected to the mud level detection telescopic rod 42, and a second end of the transverse telescopic rod 44 is connected to the protective assembly 5.

[0058] The protection assembly 5 is connected to the intelligent monitoring and rotating assembly 4. The protection assembly 5 includes a shock absorber unit 51, an anti-turbulence deflector 52, and a self-cleaning brush 53. The shock absorber unit 51 and the self-cleaning brush 53 are both located at the second end of the transverse telescopic rod 44 and are located within the space formed by the anti-turbulence deflector 52.

[0059] Specifically, in the precise mud level measurement device, the box 2 and the control component 3 are suspended above the inspection well through a foldable base 31, and the box 2 and the control component 3 include a power supply 32, a multi-stage nested electric telescopic rod group 33, and a water level ultrasonic survey probe 34.

[0060] When the mud level precise measuring device is in a non-working state, the multi-stage nested electric telescopic rod group 33 and the water level ultrasonic survey probe 34 are both in a retracted state and placed in the box 2 .

[0061] When the precise mud level measurement device enters operation, it extends the multi-stage nested electric telescopic rod assembly 33 out of the housing 2 and into the inspection well for detection. The multi-stage nested electric telescopic rod may include a well depth detection telescopic rod 41 and a mud level detection telescopic rod 42. A rotary motor 43 is also provided at the end of the mud level detection telescopic rod 42. A transverse telescopic rod 44 is provided laterally of the rotary motor 43. The transverse telescopic rod 44 is capable of horizontal expansion and contraction. A variable frequency ultrasonic detector 45 and a laser detector 46 are provided at the other end of the transverse telescopic rod 44. Both the variable frequency ultrasonic detector 45 and the laser detector 46 are located below the transverse telescopic rod 44 to facilitate detection of the liquid level below. A protective assembly 5 is also provided at the other end of the transverse telescopic rod 44.

[0062] Therefore, it can be seen that the portable telescopic design structure adopted in this application solves the problem of insufficient spatial adaptability of mechanical measuring devices caused by unstructured well chamber environments; and this structure is highly portable, the equipment is small in size, light in weight, and easy to deploy, suitable for single-person operation in complex field environments; at the same time, the device has a high degree of automation, and the equipment start and stop, telescoping, rotation, detection, data processing and early warning can all be completed by one-button control or remote wireless control, which greatly reduces the complexity of operation and the labor intensity of personnel. At the same time, in the protective component of the measuring device, by installing the shock-absorbing unit between the transverse telescopic rod and the variable frequency ultrasonic detector and the laser detector, the multi-modal probe distance value is avoided from jumping, thereby avoiding obvious errors in mud level measurement; at the same time, the setting of the anti-turbulence guide plate can effectively eliminate the vortex structure in the measurement area; and the tapered structure of the anti-turbulence guide plate reduces the disturbance of turbulence on the mud-water interface and reduces the thickness of the mud level fuzzy band. In addition, the self-cleaning brush can effectively improve the efficiency of removing pollutants from the optical window of the sensor; ultimately solving the problem of signal distortion caused by multiphase flow interference under dynamic hydraulic conditions.

[0063] Please continue reading Figures 1 to 3 .

[0064] The basic data acquisition module 11 is used to acquire basic well condition data, including but not limited to well height data and water level data.

[0065] Specifically, based on the selected scenario, determine the physical quantities to be measured, such as well depth and water level data. Then, select appropriate sensors and other acquisition equipment based on the measurement requirements, ensuring they can stably reflect the basic information for the selected scenario. Determine the sensor installation locations and then install the sensors. After installation, debug and calibrate the sensors (e.g., sensor sensitivity and measurement range) to ensure the accuracy of the measured data. After debugging and calibration, prepare the probes based on the measurement requirements (e.g., probe placement, quantity, and connection method).

[0066] The mud level measurement module 12 is used to measure the mud level data of the inspection well based on the basic well condition data. This mud level data includes bottom mud level data of the well bottom sediment and mud level data of multiple well nodes in the floc layer within the well using a multimodal sensor. The application of this module in inspection wells is mainly to meet the key needs of sludge management.

[0067] Specifically, in this embodiment, a multimodal sensor is preferably used. This means that a high-frequency ultrasonic or laser sensor penetrates the flocculent layer to directly measure the actual height of the solid sludge in the sedimentary layer. Furthermore, a multi-band signal analysis method (e.g., a 5MHz and 2MHz dual-frequency probe) can be used to distinguish the dielectric constant differences between suspended flocs and settled sludge, dynamically track the expansion potential of the floc layer, and identify the mud level data at multiple well nodes within the floc layer.

[0068] The data processing module 13 is used to fuse the mud level data of multiple well nodes measured by the multimodal sensor to obtain mud level fusion data of the floc layer in the well; the bottom hole mud level data and the mud level fusion data constitute the inspection well mud level data.

[0069] In this embodiment, the data processing module 13 is used to filter, eliminate anomalies, and normalize the collected signals, and dynamically adjust the data weights using exponentially weighted moving average and coefficient of variation analysis methods to improve data stability and accuracy.

[0070] Specifically, a data set is constructed based on mud level measurement data. The collected data is preprocessed based on the data set, and the weights of multimodal sensors (such as high-frequency ultrasonic probes and laser detectors) are dynamically adjusted using exponential weighting and coefficient of variation analysis. The final mud level data value of the floc-containing layer is calculated based on the multimodal sensor weights to improve data stability and accuracy. The measured data and historical comparison data are then stored in the cloud.

[0071] To sum up, the inspection well mud level detection system based on multimodal sensing fusion provided by this application can automatically identify water level, mud level and well height, adapt to variable hydraulic conditions, and realize mud level detection at multiple key positions in the inspection well. It has the advantages of portability, high precision, anti-interference and reusability; at the same time, it can effectively reduce the interference of disturbances, bubbles, impurities, etc. in the well on the detection signal through multi-source detection and dynamic weighting algorithm, thereby improving the stability and accuracy of mud level and floc layer thickness measurement.

[0072] Please continue reading Figure 2 and Figure 3 .

[0073] The inspection well mud level detection system 1 based on multimodal perception fusion includes: a basic data acquisition module 11, a mud level measurement module 12 and a data processing module 13. The basic data acquisition module 11 includes: a well height measurement unit 111, a water level measurement unit 112 and a measurement preparation unit 113.

[0074] The well height measurement unit 111 includes an electric well height detection telescopic rod, which is controlled to move downward to contact the well bottom to obtain the well height data of the inspection well.

[0075] Specifically, a measuring device is installed at the wellhead of the inspection well to ensure that it can stably and vertically descend to the well bottom. An electric telescopic rod is then activated and slowly lowered to the well bottom. During this process, a measuring sensor records the extension of the rod or the distance between the well bottom and the wellhead in real time. When the rod touches the well bottom, the sensor sends a signal to stop the descent. This process collects data such as the well height, measurement time, and equipment status.

[0076] It should be noted that in actual operations, appropriate measurement methods and equipment should be selected based on the specific circumstances, and relevant safety standards for operation and measurement should be followed. For underground elevation measurement, methods such as leveling and trigonometric height measurement can also be used. These methods may require the use of measurement equipment such as levels and theodolites.

[0077] The water level measurement unit 112 includes a water level detection sensor installed at the wellhead of the inspection well, which is used to measure the vertical distance from the wellhead to the water surface and obtain water level data based on the vertical distance and the well height data.

[0078] Specifically, an ultrasonic sensor is first installed at a selected location on the wellhead, enabling it to transmit ultrasonic waves vertically downward and receive reflected signals. A proper connection is ensured between the sensor and data processing equipment (e.g., a microcontroller or computer) to enable real-time data transmission. The ultrasonic sensor is then activated, and the ultrasonic signal propagates through the well. When it encounters the water surface, it reflects and is received by the sensor. The sensor records the time difference between transmission and reception of the ultrasonic wave and calculates the vertical distance to the water surface using known formulas for sound speed and distance. The water level is then calculated based on the ultrasonically measured vertical distance to the water surface and the well height.

[0079] It's important to note that the effective measurement range of an ultrasonic sensor depends on the frequency and power of its transmitter, as well as the shape and size of the well. When selecting a sensor, consider the specific needs and the potential impact of environmental factors (such as temperature and pressure) on measurement results. Additionally, ensure that there are no obstacles between the sensor and the water surface that could interfere with ultrasonic transmission.

[0080] The measurement preparation unit 113 is used to control the measurement position of the mud level measurement module according to the water level data.

[0081] In this embodiment, a preset position below the liquid surface is established based on the measurement requirements and the wellbore environment. This position should be away from areas with significant water surface fluctuations to ensure measurement stability. Based on the measured water level, a control system (e.g., a microcontroller, PLC, etc.) controls the probe's lifting mechanism (e.g., an electric telescopic rod, hydraulic cylinder, etc.) to lower the probe to the preset position below the liquid surface. After the probe reaches the preset position, the fluid environment at that location is checked for stability. Once this is confirmed, the mud level and floc layer measurement process is initiated. Therefore, the measurement preparation unit 113 controls the probe's descent to the preset position below the liquid surface based on the measured water level, ensuring a stable, non-disturbanced detection environment and ensuring the accuracy of subsequent mud level and floc layer measurements.

[0082] Please continue reading Figures 1 to 3 .

[0083] The mud level measurement module 12 includes a sediment sludge measurement unit 121 , a floc layer measurement unit 122 and a result output unit 123 .

[0084] The sediment silt measurement unit 121 includes a mud level detection telescopic rod and a variable frequency ultrasonic detector. The mud level detection telescopic rod drives the variable frequency ultrasonic detector to move up and down by extending and retracting. The variable frequency ultrasonic detector uses low-frequency ultrasound to detect the position of the sediment at the bottom of the well and obtain the mud level data at the bottom of the well.

[0085] Specifically, first, select an appropriate low-frequency ultrasonic detector to ensure its frequency can penetrate the fluid and floc layer while avoiding excessive attenuation. Then, determine the detector's measurement position based on the wellbore environment and measurement requirements to ensure coverage of the entire measurement area. Next, activate the detector and emit a low-frequency ultrasonic signal. The signal propagates through the wellbore, penetrating the fluid and floc layer and reflecting when it encounters the sludge layer. The distance between the sludge layer and the detector is then calculated based on the known sound speed and reflection time. Furthermore, the effects of the liquid and floc layer on the ultrasonic signal must be considered, and the measurement data must be corrected to determine the true sludge layer location. Repeated measurements and data analysis can improve measurement accuracy and reliability. Finally, the height of the deposited sludge is calculated based on the sludge layer position and the well bottom height.

[0086] It should be noted that low-frequency ultrasonic detection technology has the advantages of strong penetration, wide measurement range, and little environmental interference, and is suitable for the measurement of sedimentary silt.

[0087] The floc layer measurement unit 122 includes a rotating motor 43, a transverse telescopic rod 44 and a laser detector installed at the bottom end of the transverse telescopic rod 44; the variable frequency ultrasonic detector 45 and the laser detector 46 form the multimodal sensor; the first end of the transverse telescopic rod 44 is connected to the mud level detection telescopic rod 42, and the second end is installed with the variable frequency ultrasonic detector 45 and the laser detector 46, and the rotating motor 43 drives the horizontal extension and contraction of the transverse telescopic rod 44; the mud level data of multiple well nodes of the floc layer in the well are measured through the high-frequency ultrasonic detection of the variable frequency ultrasonic detector 45 and the laser detector 46.

[0088] In this embodiment, a protective component 5 is provided outside the variable-frequency ultrasonic detector 45 and the laser detector 46; the protective component 5 includes: a shock-absorbing unit 51, an anti-turbulence deflector 52 and a self-cleaning brush 53; wherein, the shock-absorbing unit 51 and the self-cleaning brush 53 are arranged at the second end of the transverse telescopic rod 44; the shock-absorbing unit 51 and the self-cleaning brush 53 are arranged in the space formed by the anti-turbulence deflector 52.

[0089] Specifically, the shock absorber unit 51 is positioned between the transverse telescopic rod 44 and the variable-frequency ultrasonic detector 45 and laser detector 46, respectively. A self-cleaning brush 53 is fixedly attached to the underside of the transverse telescopic rod 44 at one end and equipped with a cleaning brush at the other end. The anti-turbulence deflector 52, with a trapezoidal cross-section, is positioned below the end of the transverse telescopic rod 44. This deflector guides the flow of fluids (e.g., air or liquid) to prevent flow separation and vortex formation, thereby reducing turbulence intensity and improving flow stability.

[0090] Specifically, a high-frequency ultrasonic detector is used to penetrate the liquid and floc layer, measuring the upper and lower boundaries of the floc layer; a laser detector is used to provide high-precision mud level measurement to supplement the ultrasonic measurement data. Therefore, the collaborative operation of multi-modal detectors, combining the measurement data of the two probes, improves the accuracy and reliability of the measurement.

[0091] For example, the high-frequency ultrasonic and laser detectors are activated simultaneously. The probes' measurement positions and angles are adjusted based on measurement requirements to ensure coverage of the entire measurement area. Then, through a control system or server, the rotary motor's lateral telescopic mechanism is controlled to rotate within the well. This allows the probes to measure mud level datasets at various locations, such as the water inlet, outlet, well wall edge, and center. The rotary motor's rotational speed and telescopic mechanism's travel distance are adjusted based on measurement requirements. Measurement parameters are then adjusted based on changes in the well environment (such as liquid density and floc thickness) to ensure accuracy.

[0092] The result output unit 123 is configured to transmit the acquired bottom hole mud level data and the mud level data of multiple well nodes in the floc layer to the data processing module 13 .

[0093] The data processing module 13 includes: a data set construction unit 131 , a data processing unit 132 and a data storage unit 133 .

[0094] The data set construction unit 131 is configured to construct a data set based on the mud level data of the plurality of well nodes in the floc layer.

[0095] In this embodiment, the actual mud level and floc layer thickness data are obtained from the mud level measurement module to construct a data set including the actual mud level and floc layer thickness.

[0096] Specifically, real-time or periodic mud level data is first obtained from mud level measurement data, including the actual mud level (i.e., the top location of the deposited sludge) and floc thickness data obtained through detection. The collected raw data is then cleaned to remove outliers, duplicates, and invalid data. Interpolation or smoothing can also be performed to fill missing values ​​or reduce noise. The preprocessed data is then organized into a structured format. Each data point includes information such as the timestamp (or measurement time), the actual mud level, and floc thickness, including location information such as the inlet, outlet, well wall edge, and center. The constructed dataset is then verified and stored for future use.

[0097] The data processing unit 132 performs fusion processing on the mud level data of multiple well nodes measured by the multimodal sensor in the data set by smoothing the data set and configuring weight coefficients for the mud level data collected by the variable frequency ultrasonic detector and the laser detector in the data set.

[0098] In this embodiment, the data processing unit 132 is used to filter, eliminate anomalies, and normalize the collected signals, and use EWMA (Exponentially Weighted Moving-Average) and CV coefficient (Coefficient of Variation) analysis methods to dynamically adjust data weights to improve data stability and accuracy.

[0099] Specifically, the smoothing process for the data set includes smoothing the continuously collected mud level data of the floc layer by using an exponentially weighted moving average calculation method.

[0100] The calculation formula of the exponentially weighted moving average is:

[0101]

[0102] in, express The smoothed mud level data value at the moment; express The mud level data value of the floc layer at the moment; Indicates time The smoothed mud level data value; It represents the smoothing coefficient, with a value range of (0, 1). It is used to control the weight of new and old data. It is used to control the weight of new and old data. The preferred value is between 0.2 and 0.5.

[0103] In this embodiment, the coefficient of variation analysis is used to evaluate the degree of discreteness of a set of measurement data and serve as the basis for dynamic weighting; by calculating the coefficient of variation CV, the system performs real-time evaluation of the volatility of the measurement data; based on the coefficient of variation, the weights of the laser and high-frequency ultrasonic probes are allocated, and the final floc-containing layer mud level data value is calculated.

[0104] Specifically, the coefficients of variation of the mud level data collected by the variable frequency ultrasonic detector and the laser detector are calculated respectively; and weight coefficients are configured for the mud level data collected by the variable frequency ultrasonic detector and the laser detector based on the coefficients of variation and a preset variation threshold.

[0105] Specifically, the calculation formula of the coefficient of variation is:

[0106] ,

[0107] ,

[0108] ,

[0109] in, Represents the standard deviation of the data set; represents the mean of the data set; Indicates the measurements; Indicates the number of measurement samples.

[0110] By calculating the coefficient of variation (CV), the system can assess the volatility of measurement data in real time. Specifically, when the CV value is less than the variation threshold (or the CV value is small), the data is stable and reliable, and a high numerical weight is assigned to the mud level data. When the CV value is greater than the variation threshold (or the CV value is large), the data is highly volatile, and a low numerical weight is assigned to the mud level data to reduce the impact of abnormal data on the final judgment result.

[0111] Furthermore, weights of the laser and high-frequency ultrasonic probes are allocated based on the coefficient of variation, and the final mud level data value of the floc-containing layer is calculated.

[0112] The calculation formula for the weight of the mud level data configured for the variable frequency ultrasonic detector is:

[0113]

[0114] The calculation formula for the weight of the mud level data configured for the laser detector is:

[0115]

[0116] The calculation formula for fusing the mud level data of multiple well nodes measured by the multimodal sensor in the data set is:

[0117]

[0118] in, Indicates the mud level fusion data of the floc layer; Represents the weight value of the mud level data of the variable frequency ultrasonic detector; Indicates the weight value of the mud level data of the laser detector; Represents the smoothed value of the mud level data of the variable frequency ultrasonic detector; Indicates the smoothed value of the mud level data of the laser detector; Indicates the coefficient of variation of the mud level data of the variable frequency ultrasonic detector; Represents the coefficient of variation of the mud level data of the laser detector.

[0119] Through the aforementioned process, this application overcomes the physical limitations of traditional single-mode acoustic / optical sensing. By employing a combined detection technology combining multi-band acoustic impedance spectroscopy (5-200kHz) and polarized laser scattering (dual wavelengths of 1064nm / 1550nm), this method achieves layered resolution of floc layers with a density gradient of 1.02-1.15g / cm³, improving interface resolution to ≤2cm and successfully reducing the transition layer identification error from the traditional ±5cm method to ±1.2cm. Furthermore, this application employs an adaptive weighted fusion algorithm for multimodal sensor data, effectively improving the measurement system's interference reduction index and overcoming the issue of ultrasonic wave amplitude attenuation in the transition layer. Furthermore, this embodiment employs a signal distortion compensation mechanism to effectively overcome the issue of ultrasonic wave amplitude attenuation in the transition layer, maintaining effective echo resolution even in environments with an attenuation of 40-60dB. Therefore, this application effectively addresses the technical issue of interface blurring caused by variations in the optical / acoustic properties of the floc transition layer.

[0120] The data storage unit 133 is used to store real-time measurement data, local processing results and historical comparison records in a built-in storage module or a cloud database to support subsequent trend analysis and operation and maintenance decisions.

[0121] In summary, the inspection well mud level detection system based on multimodal sensing fusion described in this application, by introducing multimodal sensing fusion and a dynamic weighting algorithm, breaks through the limitations of existing single-point, static, single-physical field measurements, achieving high-precision (target error ≤±1 cm) mud level monitoring in all working conditions (including turbulent flow and irregular-shaped wells). This application can automatically identify water level, mud level, and well height, adapt to variable hydraulic conditions, and achieve mud level detection at multiple key locations within the inspection well. It has the advantages of portability, high precision, anti-interference, and reusability. Through multi-source detection and a dynamic weighting algorithm, it can effectively suppress interference with the detection signal caused by disturbances, bubbles, impurities, etc. within the well, significantly improving the stability and accuracy of mud level and floc thickness measurements.

[0122] It should be understood that the division of the modules in the above system is merely a division of logical functions. In actual implementation, they may be fully or partially integrated into a single physical entity or physically separated. Furthermore, these modules may be implemented entirely in software called by a processing element, or entirely in hardware. Alternatively, some modules may be implemented in software called by a processing element, while others may be implemented in hardware. For example, module x may be a separate processing element, or integrated into a chip in the above system. Furthermore, it may be stored in the form of program code in the memory of the above system, called by a processing element in the system to perform the functions of module x. The implementation of other modules is similar. Furthermore, these modules may be fully or partially integrated or implemented independently. The processing element described herein may be an integrated circuit with signal processing capabilities. During implementation, the steps of the above method or the modules above may be performed by hardware integrated logic circuits in the processor element or by software instructions.

[0123] See also Figure 8 and Figure 9 , respectively showing a flow chart of the mud level detection method for a manhole based on multimodal sensing fusion in one embodiment of the present application and a flow chart of the operation of the mud level detection method for a manhole based on multimodal sensing fusion in one embodiment of the present application. Figure 8 and Figure 9 As shown, this embodiment provides a method for detecting mud level in an inspection well based on multimodal perception fusion. The method comprises the following steps:

[0124] S1, collect basic data of well conditions;

[0125] S2, measuring mud level data of the inspection well based on the basic well condition data, wherein the mud level data of the inspection well includes bottom mud level data of the bottom mud deposited at the well and mud level data of multiple well nodes of the floc layer in the well measured by a multimodal sensor;

[0126] S3, fusing the mud level data of multiple well nodes measured by the multimodal sensor to obtain mud level fusion data of the floc layer in the well; the bottom well mud level data and the mud level fusion data constitute the inspection well mud level data.

[0127] Specifically, first, if the well height is unknown, the motor controls the telescopic rod a to extend downward, and when it reaches the bottom, the length of the telescopic rod a is recorded. That is the well height, retract the telescopic rod a.

[0128] Then, the ultrasonic probe A collects the ultrasonic signal data and records , which is the distance from the box to the water surface.

[0129] Then, the motor controls the telescopic rod b to move downward. The end is placed below the water surface. Probes B and C continuously collect mud level data for 1 minute. The high-frequency ultrasonic and laser detector detection data are collated to form a mud level data set. and , low-frequency ultrasonic detection data forms a real bottom-hole mud level data set .

[0130] Then, the exponentially weighted moving average algorithm is used to smooth the continuous observations, giving more weight to the recent data. The calculation formula is as follows: ,in, For the moment The smoothed value of For the moment The mud level data value of the floc-containing layer, For the moment The smoothed value of is a smoothing coefficient with a value range of (0,1), which is used to control the weight of new and old data, preferably between 0.2 and 0.5.

[0131] Calculate the coefficient of variation of a data set , according to the dynamic weighting of stability, the current mud position value of the flocculent layer is synthesized to ensure the logic of stability weight distribution.

[0132] Then, the rotary motor is controlled to rotate the probe to the water inlet, water outlet, well wall edge, and well center in turn. A limit protection mechanism is set in the system to avoid collision, and ultrasonic assisted ranging is used to determine whether there are obstacles blocking the detection path. The above steps are repeated to obtain the mud level data of each point. At the same time, the low-frequency ultrasonic detection data is Carry out processing to determine the actual mud position and important water conservancy nodes.

[0133] Finally, the data is stored locally and synchronously uploaded to the operation and maintenance system for historical trend analysis and operation and maintenance decision-making.

[0134] The following explanation is made by taking the sewage pipe of a certain intercepting combined sewer system in the central urban area of ​​a certain city as an example.

[0135] Sewage pipe A in a certain intercepting combined sewer system in the central urban area of ​​a certain city. The diameter of the sewage pipe is about 1m, and the slope is about 0.001; the inspection well is about 3.2m deep and 1.5m in diameter. The box is made of ABS engineering plastic, and the base is equipped with rubber adsorption feet to enhance stability. The telescopic rods a and b are made of carbon fiber material and are equipped with IP68 protection grade sealing rings. The control system has a built-in lithium battery, a single-chip microcomputer, and a wireless data transmission module. Probe A uses a 5MHz ultrasonic transducer, and probe B is a 650nm laser triangulation ranging module. The rotary motor uses stepper control and is equipped with a 360° angle encoder to achieve precise positioning. The system software uses a sliding window and self-learning threshold correction mechanism to further adapt to flow disturbances.

[0136] The specific implementation steps in this embodiment are as follows:

[0137] The first step is to select a typical inspection well in a city's drainage network as the test object. The inspection well is about 3 meters deep and the water depth is about 1.7 meters. There is obvious sludge deposition and floc layer distribution. Because the well depth is unknown on site, the system control module is activated to control the electric telescopic rod A to slowly descend until the bottom of the probe touches the bottom of the well and stops moving. The extended length of the telescopic rod A at this time is recorded. It is 3.18m, which is the height of the current inspection well.

[0138] In the second step, the ultrasonic probe A in the water level ultrasonic measurement unit sends a detection wave vertically downward to obtain the distance from the box to the water surface. is 1.45m, and the water level is calculated based on the well height. = 3.18m - 1.45m = 1.73m.

[0139] In the third step, the electric telescopic rod B controls the intelligent detection component to move down to a stable position of 1.60m (about 13cm below the water surface), and starts measuring the mud level data of the floc layer at this depth. The high-frequency ultrasonic probe and laser detector respectively collect the mud level data set of the floc layer. 、 , the sampling time is 1 minute, and the low-frequency ultrasonic probe collects the real mud level data set .

[0140] The fourth step is to collect continuous data from high-frequency ultrasound and laser and Applying Exponentially Weighted Moving Average (EWMA) processing, smoothing coefficient Set to 0.3 to generate smoothed mud level data and , reducing the impact of short-term disturbances on data stability.

[0141] Step 5: Calculate and Coefficient of variation of a data series and ,in, =0.12, =0.18, and the weight is calculated according to the following formula:

[0142]

[0143]

[0144] Based on this fusion, the final mud level value of the floc-containing layer is calculated :

[0145]

[0146] The sixth step is to control the rotary motor to drive the horizontal telescopic rod to rotate to the water inlet, water outlet, well wall edge and well center in sequence, repeat the detection steps to obtain mud level data at different points, and at the same time perform low-frequency ultrasonic data After processing, it was determined that the actual mud level was 1.68m.

[0147] In the seventh step, the system will synchronously upload the collected raw data, smoothed data, fusion results, CV analysis results and dredging warning logs to the cloud platform to support remote trend analysis and subsequent operation and maintenance management decisions.

[0148] From the above, it can be seen that the inspection well mud level detection method based on multimodal perception fusion described in this application can intelligently adjust the detection mode according to different inspection well sizes, water levels, and liquid characteristics, with good compatibility and strong adaptability; and can accurately measure the sludge height of drainage pipes, reduce manual judgment errors, and improve the maintenance efficiency of the drainage system.

[0149] In summary, the present application provides a system and method for detecting mud level in manholes based on multimodal sensing fusion, which has the following beneficial effects:

[0150] The inspection well mud level detection system based on multimodal sensing fusion provided by the present application can automatically identify the water level, mud level and well height, adapt to variable hydraulic conditions, and realize mud level detection at multiple key positions in the inspection well. It has the advantages of portability, high precision, anti-interference and reusability. It can effectively suppress the interference of disturbances, bubbles, impurities, etc. in the well on the detection signal through multi-source detection and dynamic weighting algorithm, and significantly improve the stability and accuracy of mud level and floc layer thickness measurement. At the same time, the inspection well mud level detection method based on multimodal sensing fusion provided by the present application can intelligently adjust the detection mode according to different inspection well sizes, water levels, and liquid characteristics. It has good compatibility and strong adaptability. It can also comply with national or industry standard trends, accurately measure the sludge height of drainage pipes, reduce manual judgment errors, and improve the maintenance efficiency of drainage systems.

[0151] The descriptions of the processes or structures corresponding to the above figures have different emphases. For parts that are not described in detail in a certain process or structure, please refer to the relevant descriptions of other processes or structures.

[0152] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Anyone skilled in the art may modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by one of ordinary skill in the art without departing from the spirit and technical concepts disclosed in this application shall be covered by the claims of this application.

Claims

1. A manhole mud level detection system based on multimodal perception fusion, characterized in that: include: A basic data acquisition module, used to collect basic well condition data; a mud level measurement module for measuring mud level data of the inspection well based on the well condition basic data, wherein the mud level data of the inspection well includes bottom mud level data of bottom mud deposited at the well bottom and mud level data of multiple well nodes of the floc layer in the well measured by a multimodal sensor; the mud level measurement module includes: a sedimentation mud measurement unit, a floc layer measurement unit, and a result output unit; The sediment silt measurement unit includes a mud level detection telescopic rod and a variable frequency ultrasonic detector. The mud level detection telescopic rod drives the variable frequency ultrasonic detector to move up and down by extending and retracting. The variable frequency ultrasonic detector uses low-frequency ultrasound to detect the position of the sediment at the bottom of the well and obtain the mud level data at the bottom of the well. The floc layer measurement unit includes a rotating motor, a transverse telescopic rod, and a laser detector mounted at the bottom end of the transverse telescopic rod; the variable-frequency ultrasonic detector and the laser detector form the multimodal sensor; the first end of the transverse telescopic rod is connected to the mud level detection telescopic rod, and the second end is mounted with the variable-frequency ultrasonic detector and the laser detector; the rotating motor drives the horizontal extension and contraction of the transverse telescopic rod; the high-frequency ultrasonic detection of the variable-frequency ultrasonic detector and the laser detector are used to measure mud level data at multiple well location nodes in the floc layer within the well; The result output unit is used to transmit the acquired bottom hole mud level data and the mud level data of multiple well nodes of the floc layer to the data processing module; A data processing module is used to fuse the mud level data of multiple well nodes measured by the multimodal sensor to obtain mud level fusion data of the floc layer in the well; the bottom mud level data and the mud level fusion data constitute the inspection well mud level data; the data processing module includes: a data set construction unit, a data processing unit, and a data storage unit; wherein, The data set construction unit constructs a data set based on the mud level data of multiple well nodes in the floc layer; The data processing unit performs fusion processing on the mud level data of multiple well nodes measured by the multimodal sensor in the data set by smoothing the data set and configuring weight coefficients for the mud level data collected by the variable frequency ultrasonic detector and the laser detector in the data set.

2. The inspection well mud level detection system based on multimodal perception fusion according to claim 1 is characterized in that: The basic well condition data includes well height data and water level data; the basic data acquisition module includes: a well height measurement unit, a water level measurement unit and a measurement preparation unit; The well height measurement unit includes an electric well height detection telescopic rod, which is controlled to move downward to contact the well bottom to obtain the well height data of the inspection well; The water level measurement unit includes a water level detection sensor installed at the wellhead of the inspection well, which is used to measure the vertical distance from the wellhead to the water surface and obtain water level data based on the vertical distance and the well height data; The measurement preparation unit is used to control the measurement position of the mud level measurement module according to the water level data.

3. The inspection well mud level detection system based on multimodal perception fusion according to claim 1 is characterized in that: The basic data acquisition module, the mud level measurement module and the data processing module are installed in a box; the box is arranged above the inspection well, and the basic data acquisition module and the mud level measurement module can be folded and retracted in the box, and can be extended from the bottom of the box to perform data acquisition and measurement; A protective component is provided on the outside of the variable-frequency ultrasonic detector and the laser detector; the protective component includes: a shock-absorbing unit, an anti-turbulence deflector and a self-cleaning brush; wherein, the shock-absorbing unit and the self-cleaning brush are arranged at the second end of the transverse telescopic rod; the shock-absorbing unit and the self-cleaning brush are arranged in the space formed by the anti-turbulence deflector.

4. The inspection well mud level detection system based on multimodal perception fusion according to claim 1 is characterized in that: Smoothing the data set includes smoothing the continuously collected mud level data of the floc layer by an exponentially weighted moving average calculation method; wherein the calculation formula of the exponentially weighted moving average calculation method is: S t =α×M t +(1-a)×S t-1 Among them, S t represents the smoothed mud level data value at time t; M t S represents the mud level data value of the floc layer at time t; t-1 represents the smoothed mud level data value at time t-1; α represents the smoothing coefficient, which ranges from (0, 1) and is used to control the data weight.

5. The inspection well mud level detection system based on multimodal perception fusion according to claim 1 is characterized in that: The configuration of weight coefficients for the mud level data collected by the variable frequency ultrasonic detector and the laser detector in the data set includes: Calculating the coefficient of variation of the mud level data collected by the variable frequency ultrasonic detector and the laser detector respectively; A weight coefficient is configured for the mud level data collected by the variable-frequency ultrasonic detector and the laser detector based on each of the variation coefficients and a preset variation threshold.

6. The inspection well mud level detection system based on multimodal perception fusion according to claim 5 is characterized in that: The calculation formula of the coefficient of variation is: Among them, σ represents the standard deviation of the data set; μ represents the mean of the data set; x i represents the i-th measurement value; n represents the number of measurement samples; When the coefficient of variation CV value is less than the variation threshold, a preset high numerical weight is configured for the mud level data; When the coefficient of variation CV value is greater than the variation threshold, a preset low numerical weight is configured for the mud level data.

7. The inspection well mud level detection system based on multimodal perception fusion according to claim 4 or 5 is characterized in that: The calculation formula for the weight of the mud level data configured for the variable frequency ultrasonic detector is: The calculation formula for the weight of the mud level data configured for the laser detector is: IN 激光 =1-W 超声 The calculation formula for fusing the mud level data of multiple well nodes measured by the multimodal sensor in the data set is: M=W 超声 ×S 超声 +W 激光 ×S 激光 Where M represents the mud position fusion data of the floc layer; W 超声 W represents the weight value of the mud level data of the variable frequency ultrasonic detector; 激光 Represents the weight value of the mud level data of the laser detector; S 超声 Represents the smoothed value of the mud level data of the variable frequency ultrasonic detector; S 激光 It represents the smoothed value of the mud level data of the laser detector; CV1 represents the coefficient of variation of the mud level data of the variable frequency ultrasonic detector; CV2 represents the coefficient of variation of the mud level data of the laser detector.

8. A method for detecting mud level in an inspection well based on multimodal perception fusion, characterized in that: The method comprises: Collect basic well condition data; Measuring mud level data of the inspection well based on the well condition basic data, wherein the mud level data of the inspection well includes bottom mud level data of bottom mud deposited at the well bottom and mud level data of multiple well nodes of the floc layer in the well measured based on a multimodal sensor; the mud level measurement module includes: a sedimentation mud measurement unit, a floc layer measurement unit, and a result output unit; The sediment silt measurement unit includes a mud level detection telescopic rod and a variable frequency ultrasonic detector. The mud level detection telescopic rod drives the variable frequency ultrasonic detector to move up and down by extending and retracting. The variable frequency ultrasonic detector uses low-frequency ultrasound to detect the position of the sediment at the bottom of the well and obtain the mud level data at the bottom of the well. The floc layer measurement unit includes a rotating motor, a transverse telescopic rod, and a laser detector mounted at the bottom end of the transverse telescopic rod; the variable-frequency ultrasonic detector and the laser detector form the multimodal sensor; the first end of the transverse telescopic rod is connected to the mud level detection telescopic rod, and the second end is mounted with the variable-frequency ultrasonic detector and the laser detector; the rotating motor drives the horizontal extension and contraction of the transverse telescopic rod; the high-frequency ultrasonic detection of the variable-frequency ultrasonic detector and the laser detector are used to measure mud level data at multiple well location nodes in the floc layer within the well; The result output unit is used to transmit the acquired bottom hole mud level data and the mud level data of multiple well nodes of the floc layer to the data processing module; The mud level data of multiple well nodes measured by the multimodal sensor are fused to obtain mud level fusion data of the floc layer in the well; the bottom mud level data and the mud level fusion data constitute the inspection well mud level data; the data processing module includes: a data set construction unit, a data processing unit, and a data storage unit; in, The data set construction unit constructs a data set based on the mud level data of multiple well nodes in the floc layer; The data processing unit performs fusion processing on the mud level data of multiple well nodes measured by the multimodal sensor in the data set by smoothing the data set and configuring weight coefficients for the mud level data collected by the variable frequency ultrasonic detector and the laser detector in the data set.

Images