Shaft construction-oriented multi-source data acquisition and fusion system and fusion method
By using an information potential field model and an adaptive sampling strategy, the problems of dynamic adjustment and spatial constraints in multi-source data acquisition and fusion during well construction were solved, achieving high-precision and stable data fusion and construction monitoring.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-12
- Publication Date
- 2026-04-07
AI Technical Summary
In existing technologies, multi-source data acquisition and fusion systems during well construction suffer from several problems, including insufficient handling of information conflicts, inability to dynamically adjust static weight models, lack of dynamic monitoring and correction of well space constraint boundaries, and fixed sensor sampling frequencies leading to insufficient information or increased redundant data in key areas.
By employing an information potential field construction module, a particle potential energy adjustment module, a constraint projection correction module, and a spatiotemporal feedback evolution module, dynamic acquisition and fusion of multi-source data are achieved through dynamic adjustment of information particles, projection correction, and adaptive sampling strategies.
It improved the accuracy of data fusion and system stability, reduced interference from abnormal data, enabled dynamic collaboration among multiple sensors, and enhanced the monitoring and decision-making capabilities during the construction process.
Smart Images
Figure CN121808698A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application belongs to the technical field of data acquisition, and particularly relates to a multi-source data acquisition and fusion system for wellbore construction. BACKGROUND
[0002] The existing multi-source data acquisition and fusion system and method for wellbore construction mainly have the following problems: In the wellbore construction process, in order to ensure construction safety and efficiency, it is usually necessary to acquire multi-source information such as drilling pressure, mud density, and formation structure in real time, and to perform fusion analysis on the data acquired by various sensors. However, the existing multi-source data acquisition and fusion system and method for wellbore construction have many technical problems, mainly including the following aspects: Insufficient conflict processing of multi-source information, when information from different sensors or data sources overlaps in space and the observation results have significant deviations, the traditional weighted average or simple filtering method cannot dynamically balance the conflicting information, resulting in abnormal fluctuations in the potential field distribution of the wellbore space. In addition, the existing technology often ignores the relative potential energy difference between multi-source information, and is easy to produce information overload or energy deviation under high coupling state. The traditional method cannot regulate the time-varying transmission characteristics of information energy, and is easy to appear response lag.
[0003] The traditional data fusion algorithm is usually static, and cannot adjust the data source weight in real time with the construction process, and cannot reflect the dynamic change characteristics in the wellbore construction process. When the formation structure, drilling pressure, mud density and other factors change with time, the static weight model cannot be adjusted in real time, resulting in disconnection between the information fusion result and the actual construction state.
[0004] The existing technology lacks dynamic monitoring and correction of the spatial constraint boundary of the wellbore, resulting in deviation of information particles from the physical constraints in space, causing local data anomalies or imbalance of information potential field. At the same time, the difference in potential field gradient between information particles lacks quantitative judgment, which cannot find local potential field anomalies in real time, nor can it implement targeted correction on abnormal areas. The abnormal information particles lack energy regulation, which causes disturbance to the overall potential field distribution, affecting the data fusion accuracy and system stability.
[0005] The sampling frequency of the sensor in the existing technology is fixed, and cannot be adjusted dynamically according to the importance of information or the local environment, resulting in insufficient information in the key area or low data acquisition efficiency. When the information weight distribution is uneven or there are abnormal areas, the fixed sampling strategy may cause key data to be missed or redundant data to increase, reducing the data fusion accuracy. The sampling and data fusion of different types of sensors in space lack a unified driving basis, and cannot realize dynamic cooperation among multiple sensors, increasing the difficulty of real-time perception in the construction process. SUMMARY
[0006] The purpose of the present application is to provide a multi-source data acquisition and fusion system for wellbore construction, which solves the problem of lack of unified driving basis for sampling and data fusion of different types of sensors in space in the prior art, and cannot realize dynamic collaboration between multiple sensors.
[0007] Another purpose of the present application is to provide a multi-source data acquisition and fusion method for wellbore construction.
[0008] The technical solution adopted by the present application is a multi-source data acquisition and fusion system for wellbore construction, comprising: An information potential field construction module acquires multi-source data of a wellbore construction site, establishes an information potential field model of a wellbore space according to the multi-source data, and maps the multi-source data into information particles carrying information weight, direction vector and confidence potential energy; A particle potential energy adjustment module dynamically adjusts the confidence potential energy of each information particle through a confidence energy conservation mechanism when different information particles overlap in space and the difference in confidence potential energy is greater than a preset confidence potential energy difference threshold; A constraint projection correction module maps the distribution of adjusted information particles to the physical boundary of the wellbore, and when the distribution of information particles deviates from the physical boundary of the wellbore, it is projected to the physical boundary of the wellbore again through the normal vector projection method; if the deviation persists, local potential field gradient correction is triggered to generate local potential field state change information; A space-time feedback evolution module tracks the evolution state of the information potential field model in the construction stage based on the local potential field state change information, records the potential field state vector and calculates the potential field change rate; when the potential field change rate exceeds a preset potential field change rate threshold, the information weight of the corresponding region is corrected; A sampling and fusion feedback module automatically adjusts the sampling strategy of each sensor according to the corrected information weight; through potential field gradient driving, self-adaptive collaborative control is realized to achieve dynamic acquisition and fusion of multi-source data in the wellbore construction process.
[0009] The present application is also characterized in that Specifically, the method for establishing an information potential field model of a wellbore space according to multi-source data comprises: Acquiring multi-source data of a wellbore construction site, the multi-source data including geological parameter data, mechanical state data and environmental state data; Performing time synchronization and space alignment processing on the multi-source data, using linear interpolation method to unify the time scale of data with different sampling frequencies, and performing space coordinate mapping and normalization processing on data of different spatial sampling regions to obtain a multi-source synchronous data set under a unified space-time reference; The feature extraction processing is performed on the multi-source synchronous data set, a principal component analysis method is used to calculate information feature components in each space unit, and a feature component matrix reflecting the space state of the wellbore is obtained; and based on the feature component matrix, the information potential energy of each position point in the wellbore space is calculated, and an information potential field model of the wellbore space is established according to the information potential energy distribution of each position point.
[0010] Specifically, the information particle acquisition method comprises: Based on the information potential energy distribution of each position point in the information potential field model of the wellbore space, the potential energy response intensity of the multi-source data at the corresponding position point is calculated, and the information weight is assigned to each data segment in the multi-source data according to the response intensity; The potential energy response intensity curve of each data segment changing with time is taken as a response feature vector, and the similarity between different data segments in the multi-source data is calculated by combining the response feature vectors of the data segments of the adjacent positions in space. The data segments with a similarity greater than a preset similarity threshold and adjacent in space are clustered to form a same multi-source fusion unit, and a unique information particle is generated for each multi-source fusion unit. Each information particle is composed of an information weight, a direction vector and a confidence potential.
[0011] Specifically, the method for dynamically adjusting the confidence potential of each information particle comprises: When different information particles in the information potential field model overlap in space and the confidence potential difference is greater than a preset confidence potential difference threshold, the potential energy response difference of each information particle relative to other overlapping information particles is calculated to form an interaction relationship between the information particles; An information energy interaction equation is constructed to dynamically adjust the confidence potential of the information particle, a smooth energy migration relationship is formed by a nonlinear Sigmoid function, high-confidence information particles absorb low-confidence information particle energy, while abnormal information particles are inhibited from interfering with the information potential field model; and the confidence potential of the information particle is updated according to the adjustment result, so that the information particle potential energy distribution tends to be consistent in space.
[0012] Specifically, the method for re-projecting to the physical boundary of the wellbore by the normal vector projection method comprises: The adjusted information particle distribution is mapped to the three-dimensional physical boundary model of the wellbore, the three-dimensional physical boundary model of the wellbore is established according to the predetermined construction design data, the three-dimensional physical boundary model of the wellbore includes the geometric information of the wellbore inner diameter, the inclination angle and the well depth, and the wellbore space constraint boundary domain is formed; It is detected whether each information particle deviates from the wellbore space constraint boundary domain, and the information particle deviating from the wellbore space constraint boundary domain is marked; for the deviated information particle, the normal vector projection method is used for repositioning, and the deviated information particle is moved to the wellbore space constraint boundary domain along the normal direction of the wellbore boundary surface.
[0013] Specifically, the method of generating local potential field state change information comprises: After the normal vector projection is completed, the position of the information particle in the wellbore space constraint boundary domain is continuously monitored; if any information particle still deviates after projection, it is determined that there is a persistent deviation phenomenon in the region, and when the persistent deviation is detected, the system triggers a local potential field gradient correction mechanism; Taking the information particle with persistent deviation as the center, a set of adjacent information particles is selected within a preset neighborhood radius range, the information particle potential field gradient within the preset neighborhood radius range is averaged and differentiated, the difference between the center information particle and the neighborhood information particle potential field gradient is calculated, and the local potential field gradient difference value is obtained; when the local potential field gradient difference value exceeds the preset local potential field gradient difference value threshold, it is determined that the local potential field is abnormal; By introducing a correction potential function, a correction force along the normal direction of the wellbore space constraint boundary domain is applied to the abnormal information particle, so that the abnormal information particle gradually converges to the wellbore space constraint boundary domain; after the correction is completed, the correction force vector applied to the abnormal information particle is taken as the local potential field state change information.
[0014] Specifically, the method of recording the potential field state vector and calculating the potential field change rate comprises: According to the local potential field state change information, the potential field state vector of each information particle in the information potential field model is periodically updated, and the potential field state vector includes the position component, the local potential field gradient vector and the correction force component of the information particle; In a continuous time interval, according to the potential field state vector of the information particle, the potential field state vector change amount of the information particle at the adjacent sampling time is calculated, and then divided by the adjacent sampling time interval to obtain the potential field change rate of the information particle.
[0015] Specifically, the method of correcting the information weight of the corresponding region comprises: When the potential field change rate of any information particle exceeds the preset potential field change rate threshold, it is determined that the potential field evolution speed of the region where the information particle is located is abnormal, and a corresponding set of information particles in the region is recorded; For each information particle in the information particle set, the information weight of the information particle in the information potential field model is adjusted according to the amplitude of the potential field change rate, and the potential field change rate is recalculated; when the potential field change rate falls within the preset potential field change rate threshold range, it is determined that the potential field of the region is restored to stability, otherwise the neighborhood analysis range is further expanded and the information weight is dynamically updated.
[0016] Specifically, the method of realizing dynamic acquisition and fusion of multi-source data in the wellbore construction process comprises: Based on the corrected information weight, the sampling strategy of each sensor is automatically adjusted, and the sampling frequency of each region in the information potential field model of the wellbore space is adaptively and cooperatively controlled through the potential field gradient driving function, so that different types of sensors collect data at a higher frequency in the region with high information weight and collect data at a lower frequency in the region with low information weight, thereby realizing dynamic acquisition and fusion of multi-source data in the wellbore construction process.
[0017] Another technical solution adopted by the present application is a multi-source data acquisition and fusion method for wellbore construction, and the steps are as follows: S1, acquire multi-source data of the wellbore construction site, establish an information potential field model of the wellbore space according to the multi-source data, and map the multi-source data into information particles carrying information weight, direction vector and confidence potential energy; S2, when different information particles overlap in space and the numerical difference is greater than the preset difference threshold, the confidence potential energy of each information particle is dynamically adjusted through the confidence energy conservation mechanism; S3, map the adjusted information particle distribution to the wellbore physical boundary, and when the information particle distribution deviates from the wellbore physical boundary, re-project it to the wellbore physical boundary through the normal vector projection method; if the deviation persists, trigger local potential field gradient correction to generate local potential field state change information; S4, based on the local potential field state change information, track the evolution state of the information potential field model in the construction stage, record the potential field state vector and calculate the potential field change rate; when the potential field change rate exceeds the preset potential field change rate threshold, correct the information weight of the corresponding region; S5, based on the corrected information weight, the sampling strategy of each sensor is automatically adjusted; through the potential field gradient driving, adaptive and cooperative control is realized, and dynamic acquisition and fusion of multi-source data in the wellbore construction process are realized.
[0018] Compared with the prior art, the present application has the following beneficial effects: An energy coupling relationship between high-confidence and low-confidence information particles is constructed through an information-energy mutual interference equation. This allows high-confidence particles to absorb and smoothly migrate the energy of low-confidence particles at the energy level, thereby forming a dynamic equilibrium state in local conflict areas. This mechanism not only adaptively adjusts the influence weights of each information source but also suppresses the mutual repulsion effect between data, making the overall distribution of the wellbore potential field more continuous and stable. A smooth energy migration relationship is constructed using a nonlinear sigmoid function, which effectively suppresses the disturbance of the overall potential field by sudden abnormal particles or signals, improving the robustness of the model. Constraining the energy migration process with a time decay parameter ensures continuous connection of data from different time periods at the energy level, achieving continuous characterization of the dynamic environment of the wellbore. The particle distribution, adjusted by confidence potential energy, tends to be spatially smooth and consistent, enabling the information potential field model to more accurately reflect the actual state changes of the wellbore and improving the accuracy of subsequent sampling control and construction monitoring.
[0019] By analyzing the potential field gradient of neighboring particles and determining thresholds, the system achieves rapid and accurate identification of abnormal regions. A correction force is applied to abnormal particles using a correction potential function to restore local potential field equilibrium, ensuring the consistency of the overall potential field distribution. Information on changes in the local potential field state provides reliable feedback, enabling the system to dynamically adjust information weights and sampling strategies, reducing interference from abnormal data and improving the accuracy and real-time performance of data fusion during the construction phase. Through rapid correction of local potential field anomalies, the system can achieve real-time monitoring, dynamic correction, and closed-loop control of multi-source information during construction, enhancing the intelligence level of the data acquisition system.
[0020] Adaptive sampling strategies and information weighting are employed to ensure high-quality data acquisition in key areas while reducing redundant sampling. The combination of information weight adjustment and potential field gradient-driven approaches effectively suppresses interference from anomalous information, resulting in more stable and accurate data fusion results. Different types of sensors dynamically adjust their sampling based on the potential field gradient within the wellbore space, achieving coordinated spatial and temporal acquisition and improving monitoring and decision-making capabilities during construction. The sampling frequency is reduced in low-weight areas, minimizing unnecessary data acquisition and transmission overhead and enhancing overall system efficiency. Attached Figure Description
[0021] Figure 1 This is a schematic diagram of the multi-source data acquisition and fusion system for well construction according to the present invention; Figure 2 This is a schematic diagram of the multi-source data acquisition and fusion method for well construction according to the present invention. Detailed Implementation
[0022] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be clearly and completely described below, obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.
[0023] Embodiment 1 The wellbore construction-oriented multi-source data acquisition and fusion system of the present application has a structure as shown in Figure 1 The wellbore construction-oriented multi-source data acquisition and fusion system of the present application has a structure as shown in The information potential field construction module acquires multi-source data of the wellbore construction site, establishes an information potential field model of the wellbore space according to the multi-source data, and maps the multi-source data into information particles carrying information weight, direction vector and confidence potential energy; The particle potential energy adjustment module dynamically adjusts the confidence potential energy of each information particle through a confidence energy conservation mechanism when different information particles overlap in space and the difference in confidence potential energy is greater than a preset confidence potential energy difference threshold; The constraint projection correction module maps the distribution of the adjusted information particles to the physical boundary of the wellbore, and when the distribution of the information particles deviates from the physical boundary of the wellbore, it is projected to the physical boundary of the wellbore again through the normal vector projection method; if the deviation persists, a local potential field gradient correction is triggered to generate local potential field state change information; The space-time feedback evolution module tracks the evolution state of the information potential field model in the construction stage based on the local potential field state change information, records the potential field state vector and calculates the potential field change rate; when the potential field change rate exceeds a preset potential field change rate threshold, the information weight of the corresponding region is corrected; The sampling fusion feedback module automatically adjusts the sampling strategy of each sensor according to the corrected information weight; through potential field gradient driving, self-adaptive cooperative control is realized to achieve dynamic acquisition and fusion of multi-source data in the wellbore construction process.
[0024] Embodiment 2 Based on embodiment 1, the method for establishing an information potential field model of the wellbore space according to multi-source data comprises: Acquiring multi-source data of the wellbore construction site, the multi-source data including geological parameter data, mechanical state data and environmental state data; The geological parameter data includes formation lithology, formation density, porosity, water content, ground stress, fault information, drilling fluid performance and gas content; the mechanical state data includes drilling parameters, downhole tool state, mud pump parameters, power consumption data, equipment temperature and wear information; the environmental state data includes wellbore wall temperature, pressure distribution, vibration signal, acoustic signal, gas concentration change data, drilling fluid level and flow rate; It should be noted that the multi-source data in the embodiment is obtained by a plurality of types of sensors arranged at the wellbore construction site, and the plurality of types of sensors include a geological parameter measuring sensor arranged in the well and a mechanical operation state sensor and an environmental state sensing sensor arranged in the ground equipment. For example, the geological parameter measuring sensor can include a gamma ray sensor, a density measuring sensor, a nuclear magnetic resonance sensor, and an ultrasonic sensor, wherein the geological parameter measuring sensor is used to obtain geological parameter data such as formation lithology, formation density, porosity, ground stress, and gas content. The mechanical operation state sensor can include a drilling pressure sensor, a rotation speed sensor, a torque sensor, a current voltage sensor, and a temperature sensor, and the mechanical operation state sensor is used to obtain mechanical state data such as drilling parameters, mud pump operation parameters, power consumption, and equipment temperature. The environmental state sensing sensor can include a downhole temperature sensor, a pressure sensor, an acceleration sensor, an acoustic sensor, and a flow sensor, and the environmental state sensing sensor is used to obtain environmental state data such as temperature, pressure, vibration, acoustic signal, and drilling fluid flow state in the wellbore. The multi-source data is subjected to time synchronization and space alignment processing, linear interpolation method is used for time scale unification of data with different sampling frequencies, and space coordinate mapping and normalization processing is performed on data with different spatial sampling regions, to obtain a multi-source synchronous data set under a unified space-time reference. The multi-source synchronous data set is subjected to feature extraction processing, principal component analysis method is used to calculate information feature components in each spatial unit, to obtain a feature component matrix reflecting the spatial state of the wellbore; and information potential of each position point in the wellbore space is calculated based on the feature component matrix, and an information potential field model of the wellbore space is established according to the information potential distribution of each position point.
[0025] The information particle acquisition method includes: Based on the information potential distribution of each position point in the information potential field model of the wellbore space, the potential energy response intensity of the multi-source data at the corresponding position point is calculated, and the information weight is assigned to each data segment in the multi-source data according to the response intensity; The curve of the potential energy response intensity of each data segment changing with time is taken as a response feature vector, and the response feature vectors of the data segments of the adjacent positions in space are combined to calculate the similarity between different data segments in the multi-source data; The data segments with a similarity greater than a preset similarity threshold and adjacent in space are clustered to form a same multi-source fusion unit, and a unique information particle is generated for each multi-source fusion unit; each information particle is composed of information weight, direction vector, and confidence potential.
[0026] The information weight is used to represent the contribution degree of the information particle to the local potential energy change in the whole potential field; the direction vector is determined according to the gradient direction of the potential field, and is used to indicate the trend of information flow and the interaction path of the adjacent space region; the confidence potential is calculated by fusing the variance, covariance and time stability of the original data, and is used to represent the confidence degree and dynamic stability of the information particle in the potential field.
[0027] Embodiment 3 On the basis of embodiment 2, the method for dynamically adjusting the confidence potential of each information particle comprises: When different information particles in the information potential field model overlap in space and the confidence potential difference is greater than a preset confidence potential difference threshold, the potential energy response difference of each information particle relative to other overlapping information particles is calculated to form the mutual interference relationship between the information particles; An information energy mutual interference equation is constructed to dynamically adjust the confidence potential of the information particle, a smooth energy migration relationship is formed by a nonlinear Sigmoid function to enable the high-confidence information particle to absorb the energy of the low-confidence information particle, while the interference of the abnormal information particle on the information potential field model is inhibited; and the confidence potential of the information particle is updated according to the adjustment result, so that the potential energy distribution of the information particle tends to be consistent in space.
[0028] The information energy mutual interference equation is: ; wherein, represents the confidence potential change amount of the information particle , and is used to describe the energy increase or decrease amplitude of the particle in the information potential field modulation process; a positive value represents potential energy enhancement, and a negative value represents potential energy attenuation; represents the information mutual interference coefficient, and is used to adjust the strength of the potential energy interaction between the information particles; represents the current confidence potential of the information particle ; represents the current confidence potential of the information particle ; represents the potential energy response difference of each information particle relative to other overlapping information particles; represents the confidence potential difference item between the information particles, and is used to measure the energy transfer direction and size: if the difference is positive, it represents that the energy migrates from the information particle to the information particle , and if it is negative, it represents that the energy of the information particle diffuses outward; represents the direction consistency factor of the information particle and the information particle in the feature space, and is used to reflect the direction similarity of the two particles in the dimension of the multi-source data feature vector; represents the time decay parameter, and is used to control the decreasing trend of the potential energy mutual interference with time; represents the absolute difference of the confidence potential between particles; and represents the index of the information particle; Embodiment 4 On the basis of embodiment 3, the method of re-projecting to the physical boundary of the wellbore by the normal vector projection method comprises: mapping the adjusted information particle distribution into the three-dimensional physical boundary of the wellbore, establishing a three-dimensional physical boundary model of the wellbore according to predetermined construction design data, the three-dimensional physical boundary model comprising geometric information of the inner diameter of the well wall, the inclination angle and the well depth, forming a wellbore space constraint boundary domain; detecting whether each information particle deviates from the wellbore space constraint boundary domain, marking the information particles deviating from the wellbore space constraint boundary domain; for the deviated information particles, repositioning is performed by the normal vector projection method, and the deviated information particles are moved to the wellbore space constraint boundary domain along the normal direction of the wellbore boundary surface.
[0029] The method of generating local potential field state change information comprises: After the normal vector projection is completed, the system continuously monitors the position of the information particles in the wellbore space constraint boundary domain; if any information particle still deviates after projection, it is determined that there is a persistent deviation in this region, and when the persistent deviation is detected, the system triggers the local potential field gradient correction mechanism; taking the information particle with persistent deviation as the center, selecting a set of adjacent information particles within a preset neighborhood radius range, performing averaging and differentiation analysis on the potential field gradient of the information particles within the preset neighborhood radius range, calculating the difference between the center information particle and the neighborhood information particle potential field gradient, and obtaining the local potential field gradient difference value; when the local potential field gradient difference value exceeds the preset local potential field gradient difference value threshold, it is determined that the local potential field is abnormal; the local potential field gradient difference value is: ; wherein, represents the local potential field gradient difference value, and represents the deviation degree of the center information particle and the neighborhood information particle in the potential field gradient; represents the local potential field gradient vector of the information particle , which contains direction and intensity information; represents the local potential field gradient vector of the neighborhood information particle ; represents the number of neighborhood information particles; represents the index of the neighborhood information particle; By introducing a correction potential function, a correction force along the normal direction of the wellbore space constraint boundary domain is applied to the abnormal information particle, so that the abnormal information particle gradually converges to the wellbore space constraint boundary domain; the correction potential function is: ; wherein, represents a correction force vector applied on the information particle ; represents a correction coefficient, adjusting the correction strength; represents a wellbore boundary normal vector, pointing to the inside of the wellbore space constraint boundary domain; after the correction is completed, the correction force vector applied on the abnormal information particle is taken as the local potential field state change information.
[0030] Embodiment 5 On the basis of embodiment 4, the method for recording the potential field state vector and calculating the potential field change rate comprises: According to the local potential field state change information, periodically update the potential field state vector of each information particle in the information potential field model, the potential field state vector comprising the position component, the local potential field gradient vector and the correction force component of the information particle, for describing the instantaneous equilibrium state of the information particle in the wellbore space constraint domain; In the continuous time interval, according to the potential field state vector of the information particle, calculate the potential field state vector change amount of the information particle at the adjacent sampling time, and divide by the adjacent sampling time interval to obtain the potential field change rate of the information particle.
[0031] The potential field change rate of the information particle is: ; wherein, represents the potential field change rate of the information particle at the time ; represents the potential field state vector of the information particle at the adjacent sampling time ; represents the potential field state vector of the information particle at the time ; represents the adjacent sampling time interval; represents the index of the time; The method for correcting the information weight of the corresponding region comprises: When it is detected that the potential field change rate of any information particle exceeds the preset potential field change rate threshold, it is determined that the potential field evolution speed of the region where the information particle is located is abnormal, and a corresponding information particle set in the region is recorded; For each information particle in the information particle set, adjust the information weight of the information particle in the information potential field model according to the amplitude of the potential field change rate, the greater the potential field change rate, the higher the weight decay amplitude of the corresponding information particle, so as to reduce the interference of the region on the overall potential field distribution; And recalculate the potential field change rate, when the potential field change rate falls within the preset potential field change rate threshold range, it is determined that the potential field of the region is restored to stability, otherwise further expand the neighborhood analysis range and dynamically update the information weight.
[0032] Embodiment 6 On the basis of embodiment 5, the method for realizing dynamic acquisition and fusion of multi-source data in the wellbore construction process comprises: Based on the corrected information weight, the sampling strategy of each sensor is automatically adjusted, specifically including: for each local region in the information potential field model of the wellbore space, the system automatically determines the sampling strategy of the sensor by analyzing the weight and local potential energy gradient of the information particles in the region; For the region with high weight and large potential energy gradient, it indicates that the information of the region changes dramatically, and the system automatically increases the sampling frequency or increases the sampling density of the sensor to enhance the spatio-temporal resolution of the data. For the region with low weight or gentle potential energy gradient, it indicates that the information of the region is stable, and the system appropriately reduces the sampling frequency to save energy consumption and computing resources. For the abnormal region, if the local potential field shows abnormality or detects heterogenous interference, the system triggers a multi-channel verification sampling strategy to improve the data reliability and accuracy through multi-sensor collaborative acquisition.
[0033] And through the potential field gradient driving function, the sampling frequency of each region in the information potential field model of the wellbore space is adaptively and cooperatively controlled. Different types of sensors collect data at a higher frequency in regions with high information weight and at a lower frequency in regions with low information weight, thereby realizing dynamic acquisition and fusion of multi-source data in the wellbore construction process.
[0034] The potential field gradient driving function is: ; wherein, represents the region sampling frequency at the spatial position and the time , that is, the actual sampling speed of the corresponding region sensor; represents the reference sampling frequency, which refers to the sampling frequency when the potential field is stable and the information particle weight is the default value; represents the sensitivity factor, which is used to adjust the response degree of the sampling frequency to the change of the potential energy gradient; represents the local potential field gradient vector of the information particle at the spatial position and the time ; represents the maximum local potential field gradient vector of the information potential field model of the wellbore space; represents the corrected information weight.
[0035] The preset confidence potential energy difference threshold is set by the staff, and the average value of multiple confidence potential energy differences is taken as the preset confidence potential energy difference threshold by collecting different preset confidence potential energy differences. Similarly, the preset potential field change rate threshold and the preset similarity threshold are set.
[0036] In this embodiment, the energy coupling relationship between high-confidence information particles and low-confidence information particles is established by the information energy mutual interference equation, so that the high-confidence particles can absorb and smoothly migrate the energy of the low-confidence particles at the energy level, thereby forming a dynamic equilibrium state in the local conflict area. This mechanism not only adaptively adjusts the influence weight of each information source, but also suppresses the mutual repulsion effect between data, making the overall distribution of the wellbore potential field more continuous and stable. The smooth energy migration relationship is constructed by using the nonlinear Sigmoid function, which can effectively suppress the disturbance of sudden abnormal particles or abnormal signals on the overall potential field and improve the robustness of the model. The energy migration process is constrained by the time decay parameter, so that the data of different time periods are continuously connected at the energy level, realizing the continuous representation of the dynamic environment of the wellbore. The particle distribution adjusted by the confidence potential tends to be smooth and consistent in space, so that the information potential field model can more accurately reflect the changes of the real state of the wellbore and improve the accuracy of subsequent sampling control and construction monitoring.
[0037] The abnormal area is quickly and accurately identified through neighborhood particle potential field gradient analysis and threshold determination. The abnormal particles are subjected to a correction force by the correction potential function, so that the local potential field restores balance and ensures the consistency of the overall potential field distribution. The local potential field state change information provides a reliable feedback basis, so that the system can dynamically adjust the information weight and sampling strategy, reduce abnormal data interference, and improve the accuracy and real-time performance of data fusion in the construction phase. Through rapid correction of the local potential field anomaly, the system can realize real-time monitoring, dynamic correction and closed-loop control of multi-source information during the construction process, improving the intelligent level of the data acquisition system.
[0038] Through adaptive sampling strategy and information weight driving, high-quality data acquisition in key areas is ensured, while redundant sampling is reduced. The combination of information weight adjustment and potential field gradient driving effectively suppresses abnormal information interference, making the data fusion result more stable and accurate. Different types of sensors dynamically adjust sampling according to the potential field gradient in the wellbore space, realizing collaborative acquisition in space and time, and improving the monitoring and decision-making ability of the construction process. The sampling frequency in the low-weight area is reduced, reducing unnecessary data acquisition and transmission overhead, and improving the overall efficiency of the system.
[0039] Embodiment 7 The part not described in detail in this embodiment is described in Embodiment 6. A multi-source data acquisition and fusion method for wellbore construction is provided, and the flow is as shown in Figure 2 The steps are as follows: S1, acquiring multi-source data of the wellbore construction site, establishing an information potential field model of the wellbore space according to the multi-source data, and mapping the multi-source data into information particles carrying information weight, direction vector and confidence potential; S2. When different information particles overlap in space and the numerical difference is greater than the preset difference threshold, the confidence potential energy of each information particle is dynamically adjusted through the confidence energy conservation mechanism. S3. Map the adjusted information particle distribution onto the physical boundary of the wellbore. When the information particle distribution deviates from the physical boundary of the wellbore, reproject it onto the physical boundary of the wellbore using the normal vector projection method. If the deviation persists, trigger the local potential field gradient correction and generate local potential field state change information. S4. Based on the local potential field state change information, track the evolution of the information potential field model during the construction phase, record the potential field state vector and calculate the potential field change rate; when the potential field change rate exceeds the preset potential field change rate threshold, correct the information weight of the corresponding region. S5. Based on the corrected information weights, automatically adjust the sampling strategies of each sensor; and achieve dynamic acquisition and fusion of multi-source data during well construction through adaptive collaborative control driven by potential field gradient.
[0040] Since the electronic device described in this embodiment is the electronic device used to implement the multi-source data acquisition and fusion system and method for wellbore construction described in this application, those skilled in the art can understand the specific implementation methods and various variations of the electronic device in this embodiment based on the multi-source data acquisition and fusion system and method for wellbore construction described in this application. Therefore, how the electronic device implements the method in this application will not be described in detail here. Any electronic device used by those skilled in the art to implement the multi-source data acquisition and fusion system and method for wellbore construction described in this application falls within the scope of protection of this application.
[0041] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.
[0042] The above description is merely a preferred embodiment of the present invention, and the scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for users of ordinary technical skills, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.
Claims
1. A multi-source data acquisition and fusion system for wellbore construction, characterized in that, include: The information potential field construction module collects multi-source data from the well construction site, establishes an information potential field model of the well space based on the multi-source data, and maps the multi-source data into information particles carrying information weights, direction vectors, and confidence potential energy. The particle potential energy adjustment module dynamically adjusts the confidence potential energy of each information particle through a confidence energy conservation mechanism when different information particles overlap in space and the confidence potential energy difference is greater than a preset confidence potential energy difference threshold. The constrained projection correction module maps the adjusted information particle distribution onto the physical boundary of the wellbore. When the information particle distribution deviates from the physical boundary of the wellbore, it is reprojected onto the physical boundary of the wellbore using the normal vector projection method. If the deviation persists, the local potential field gradient correction is triggered, generating local potential field state change information. The spatiotemporal feedback evolution module tracks the evolution of the information potential field model during the construction phase based on local potential field state change information, records the potential field state vector, and calculates the potential field change rate; when the potential field change rate exceeds the preset potential field change rate threshold, the information weight of the corresponding region is corrected. The sampling fusion feedback module automatically adjusts the sampling strategy of each sensor based on the corrected information weights. Adaptive and collaborative control driven by potential field gradient is used to achieve dynamic acquisition and fusion of multi-source data during well construction.
2. The multi-source data acquisition and fusion system for wellbore construction according to claim 1, characterized in that, The method for establishing an information potential field model of the wellbore space based on multi-source data includes: Collect multi-source data from the well construction site, including geological parameter data, mechanical condition data, and environmental condition data; Multi-source data is synchronized in time and aligned in space. Data with different sampling frequencies is unified in time scale by using linear interpolation. Data with different spatial sampling regions is mapped and normalized in space coordinates to obtain a multi-source synchronized dataset under a unified spatiotemporal reference. Feature extraction processing is performed on the multi-source synchronous dataset, and principal component analysis is used to calculate the information feature components in each spatial unit to obtain the feature component matrix reflecting the state of the wellbore space. Based on the feature component matrix, the information potential energy of each location point in the wellbore space is calculated, and the information potential field model of the wellbore space is established according to the information potential energy distribution of each location point.
3. The multi-source data acquisition and fusion system for wellbore construction according to claim 2, characterized in that, The method for acquiring the information particles includes: Based on the information potential energy distribution results of each location point in the information potential field model of the wellbore space, the potential energy response intensity of the multi-source data at the corresponding location point is calculated, and information weights are assigned to each data segment in the multi-source data according to the response intensity. The curve of the potential energy response intensity of each data segment changing with time is used as the response feature vector. Combined with the response feature vectors of data segments at spatially adjacent locations, the similarity between different data segments in multi-source data is calculated. Data segments with similarity greater than a preset similarity threshold and spatially adjacent are clustered to form the same multi-source fusion unit. A unique information particle is generated for each multi-source fusion unit. Each information particle consists of three parts: information weight, direction vector, and confidence potential energy.
4. The multi-source data acquisition and fusion system for wellbore construction according to claim 3, characterized in that, The method for dynamically adjusting the confidence potential energy of each information particle includes: When different information particles in the information potential field model overlap in space and the confidence potential energy difference is greater than the preset confidence potential energy difference threshold, the potential energy response difference of each information particle relative to other overlapping information particles is calculated to form the mutual interference relationship between information particles. An information energy mutual interference equation is constructed to dynamically adjust the confidence potential energy of information particles. A smooth energy transfer relationship is formed through a nonlinear Sigmoid function, which enables high-confidence information particles to absorb the energy of low-confidence information particles, while suppressing the interference of abnormal information particles on the information potential field model. The confidence potential energy of information particles is updated according to the adjustment results, so that the potential energy distribution of information particles tends to be consistent in space.
5. The multi-source data acquisition and fusion system for wellbore construction according to claim 4, characterized in that, The method of reprojecting onto the physical boundary of the wellbore using normal vector projection includes: The adjusted information particle distribution is mapped onto the three-dimensional physical boundary of the wellbore. A three-dimensional physical boundary model of the wellbore is established based on the predetermined construction design data. The three-dimensional physical boundary model includes the geometric information of the wellbore inner diameter, inclination angle and well depth, forming the spatial constraint boundary domain of the wellbore. Detect whether each information particle deviates from the wellbore spatial constraint boundary domain, and mark the information particles that deviate from the wellbore spatial constraint boundary domain; for the deviated information particles, use the normal vector projection method to reposition them, and move the deviated information particles into the wellbore spatial constraint boundary domain along the normal direction of the wellbore boundary surface.
6. The multi-source data acquisition and fusion system for wellbore construction according to claim 5, characterized in that, The method for generating local potential field state change information includes: After the normal vector projection is completed, the position of the information particle in the wellbore space constraint boundary domain is continuously monitored; if any information particle still deviates after projection, it is determined that there is a continuous deviation phenomenon in the region. When the continuous deviation is detected, the system triggers the local potential field gradient correction mechanism. Centered on the continuously deviating information particle, a set of neighboring information particles is selected within a preset neighborhood radius. The potential field gradient of the information particles within the preset neighborhood radius is averaged and analyzed for differences. The difference between the potential field gradient of the central information particle and the neighboring information particles is calculated to obtain the local potential field gradient difference value. When the local potential field gradient difference value exceeds the preset local potential field gradient difference value threshold, it is determined that the local potential field is abnormal. A correction potential function is introduced to apply a correction force along the normal direction of the wellbore space constraint boundary domain to the anomalous information particles, so that the anomalous information particles gradually converge to the wellbore space constraint boundary domain; after the correction is completed, the correction force vector applied to the anomalous information particles is used as the local potential field state change information.
7. A multi-source data acquisition and fusion system for wellbore construction according to claim 6, characterized in that, The method for recording the potential field state vector and calculating the potential field change rate includes: Based on the information of local potential field state changes, the potential field state vector of each information particle in the information potential field model is periodically updated. The potential field state vector includes the position component of the information particle, the local potential field gradient vector, and the correction force component. Within a continuous time interval, based on the potential state vector of the information particle, the change in the potential state vector of the information particle at adjacent sampling times is calculated, and then divided by the adjacent sampling time interval to obtain the potential change rate of the information particle.
8. A multi-source data acquisition and fusion system for wellbore construction according to claim 7, characterized in that, The method for correcting the information weights of the corresponding regions includes: When the potential field change rate of any information particle exceeds the preset potential field change rate threshold, it is determined that the potential field evolution rate of the region where the information particle is located is abnormal, and the corresponding set of information particles in the region is recorded. For each information particle in the information particle set, the information weight of the information particle in the information potential field model is adjusted according to the magnitude of the potential field change rate, and the potential field change rate is recalculated. When the potential field change rate falls back to the preset potential field change rate threshold range, it is determined that the potential field in the region has recovered to stability; otherwise, the neighborhood analysis range is further expanded and the information weight is dynamically updated.
9. The multi-source data acquisition and fusion system for wellbore construction according to claim 8, characterized in that, The method for achieving dynamic acquisition and fusion of multi-source data during well construction includes: Based on the corrected information weights, the sampling strategy is automatically adjusted, and the sampling frequency of each region in the information potential field model of the wellbore space is adaptively and collaboratively controlled through the potential field gradient driving function. Different types of sensors collect data at a higher frequency in regions with high information weights and at a lower frequency in regions with low information weights, thereby realizing the dynamic acquisition and fusion of multi-source data during the wellbore construction process.
10. A multi-source data acquisition and fusion method for wellbore construction, implemented using any one of claims 1 to 9, characterized in that, The steps are as follows: S1. Based on the multi-source data collected from the well construction site, establish an information potential field model of the well space, and map the multi-source data into information particles carrying information weights, direction vectors and confidence potential energy. S2. When different information particles overlap in space and the numerical difference is greater than the preset difference threshold, the confidence potential energy of each information particle is dynamically adjusted through the confidence energy conservation mechanism. S3. Map the adjusted information particle distribution onto the physical boundary of the wellbore. When the information particle distribution deviates from the physical boundary of the wellbore, reproject it onto the physical boundary of the wellbore using the normal vector projection method. If the deviation persists, trigger the local potential field gradient correction and generate local potential field state change information. S4. Based on the local potential field state change information, track the evolution of the information potential field model during the construction phase, record the potential field state vector and calculate the potential field change rate; when the potential field change rate exceeds the preset potential field change rate threshold, correct the information weight of the corresponding region. S5. Based on the corrected information weights, automatically adjust the sampling strategies of each sensor; and achieve dynamic acquisition and fusion of multi-source data during well construction through adaptive collaborative control driven by potential field gradient.