CN108445767B - Visual on-site safety supervision and management system based on unmanned aerial vehicle - Google Patents
Visual on-site safety supervision and management system based on unmanned aerial vehicle Download PDFInfo
- Publication number
- CN108445767B CN108445767B CN201810467918.5A CN201810467918A CN108445767B CN 108445767 B CN108445767 B CN 108445767B CN 201810467918 A CN201810467918 A CN 201810467918A CN 108445767 B CN108445767 B CN 108445767B
- Authority
- CN
- China
- Prior art keywords
- module
- aerial vehicle
- unmanned aerial
- control
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Alarm Systems (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention discloses a visual field safety supervision and management system based on an unmanned aerial vehicle, which comprises: an unmanned aerial vehicle unit: the system is used for collecting data information in real time and supervising the site safety; a server unit: the system is used for storing data acquired by the unmanned aerial vehicle unit in real time and making corresponding control instruction operation in time; a mobile terminal: the data processing system is used for assisting in storing data returned by the unmanned aerial vehicle, analyzing the data and performing real-time operation, and is connected with the server unit to realize synchronous data sharing. The visual field safety supervision and management system based on the unmanned aerial vehicle provided by the invention can realize the control of the unmanned aerial vehicle unit only by an authorizer through the mobile terminal or the server unit, is easy to operate and has high automation degree.
Description
Technical Field
The invention relates to the technical field of visual field safety supervision and management, in particular to a visual field safety supervision and management system based on an unmanned aerial vehicle.
Background
At present, six power grids are built in China, the total length of a power transmission line exceeds 115 kilometers, and the power transmission line of 500kV or more becomes the main power of power grid power transmission in each area. China's territory breadth is wide, the terrain is relatively complex, hills are more, plains are fewer, and the meteorological conditions are complex and variable, so that certain difficulty is brought to the construction of a cross-region power grid and an ultrahigh voltage transmission line project, and in addition, the maintenance and the maintenance after the construction can not meet the requirements of high efficiency and rapidness only by means of the existing inspection means and conventional tests, and a good effect can not be achieved. And the application of the unmanned aerial vehicle technology can well complete tasks such as line inspection, line erection, line planning and the like, but no specific implementation case exists in the aspect of safety control of the infrastructure construction site at present.
The construction sites of power grid infrastructure are diversified and wide, and a construction unit puts in a large amount of manpower to supervise and control the safety and quality of the construction site in order to manage the site. Because the construction site environment is complex, the personnel mobility is high, and the personnel quality level is limited, a set of construction site safety real-time control system for site safety management has been developed and constructed in Anhui power transmission and transformation at present. The system forms an effective management and control mode through the handheld terminal, the wireless ball machine and the monitoring center, and has initial effect. But the current management mode also belongs to a semi-automatic management mode.
Disclosure of Invention
The technical problem to be solved by the invention is the defect that the prior art also belongs to a semi-automatic management mode.
The invention is realized by the following technical scheme: a visual on-site safety supervision management system based on unmanned aerial vehicle, the system includes:
an unmanned aerial vehicle unit: the system is used for collecting data information in real time and supervising the site safety;
a server unit: the system is used for storing data acquired by the unmanned aerial vehicle unit in real time and making corresponding control instruction operation in time;
a mobile terminal: the data processing system is used for assisting in storing data returned by the unmanned aerial vehicle, analyzing the data and performing real-time operation, and is connected with the server unit to realize synchronous data sharing.
As one preferable aspect of the present invention, the drone unit includes: the system comprises an unmanned aerial vehicle power output module, a flight control module, a camera holder unit, a flight data acquisition module, a data storage module, a 4G wireless transmission module and a 4G network control module; the flight data acquisition module is connected with the data storage module and is used for acquiring and storing flight state data and unmanned aerial vehicle equipment parameters; the 4G network control module is connected with the holder control module and the flight control module, so that the flight of the remote control unmanned aerial vehicle is guaranteed, the flight radius of the unmanned aerial vehicle is enlarged, and management personnel can be guaranteed to control the flight state of the unmanned aerial vehicle and the camera holder state at any time and any place through the mobile terminal; the data storage module is connected with the flight data acquisition module and the video acquisition module and is used for storing data information acquired by each module in real time; the 4G wireless transmission module is connected with the data storage module and the server-side 4G wireless transmission module and transmits data back in real time; the unmanned aerial vehicle power output module is connected with each module and supplies power for each module.
As one of the preferable modes of the invention, the unmanned aerial vehicle unit further comprises a voice broadcasting module, and the voice broadcasting module is connected with the 4G network control module and is used for broadcasting background calling content and sending an alarm notification.
As one of the preferable modes of the invention, the unmanned aerial vehicle unit further comprises a flight electricity testing module, and the flight electricity testing module is connected with the data storage module and is used for automatically avoiding high-voltage live cables or objects and timely informing relevant construction personnel of keeping away from a dangerous source.
As one of the preferable modes of the invention, the camera pan-tilt unit comprises a video acquisition module and a pan-tilt control module; the video acquisition module is connected with the data storage module and is used for live broadcasting and video recording; and the holder control module is connected with the 4G network control module and is used for controlling the angle of the camera.
As one of preferable embodiments of the present invention, the server unit includes: the system comprises a 4G wireless transmission module, a data storage module, a streaming media service module, a flight control service module, an intelligent analysis service module and a server power supply output module; the server power output module is connected with other modules and supplies power to the modules; the streaming media service module is connected with the data storage module and receives real-time video data pushed by the unmanned aerial vehicle in real time; the intelligent analysis service module is connected with the streaming media service module and the data storage module, captures violation information of a site in real time and commands site personnel to modify the violation information; the flight control service module is connected with the data storage module, receives flight data returned by the unmanned aerial vehicle and stores data information; the 4G wireless transmission module is connected with the data storage module and the 4G wireless transmission module at the unmanned aerial vehicle end, so that real-time data communication and storage are realized.
As one of the preferable modes of the invention, the mode of the intelligent analysis service module capturing the field violation information in real time is specifically that the basic target unit in the scene is detected through a target detection algorithm based on deep learning; the detection algorithm comprises the following steps:
(1) dividing the image into S-S grids, and when the center of one target falls into a certain grid, the grid is responsible for detecting the corresponding target;
(2) predicting B bounding boxes in each grid, and confidence scores of the bounding boxes, wherein when predicting each bounding box by using the YOLO model, the confidence value of the bounding box is predicted, which can be defined as
The value indicates the reliability of the bounding box containing the object and the reliability of the bounding box;
(3) when the model is trained, if the corresponding grid does not contain the target, the confidence value is equal to 0, otherwise, the confidence value is equal to the predicted box and the IOU of the ground route, and each bounding box contains 5 values: x, y, w, h, confidence; wherein, x and y represent the center of the bounding box, w and h represent the width and height of the bounding box, and confidence represents the IOU of the bounding box and the ground route; the confidence value is predicted and not actually calculated, and each grid predicts C conditional probabilities Pr (Class)i|Object)；
(4) During testing, multiplying the confidence coefficient of the bounding box with the class probability to obtain a confidence score of a specific class; the confidence score represents the probability that the class appears in the bounding box and also represents the degree to which the bounding box matches the target; the calculation formula is as follows:
as one of the preferable modes of the invention, in the mode that the intelligent analysis service module captures the site violation information in real time, after the basic target unit in the scene is detected by the target detection algorithm based on deep learning, the safety helmet and the dangerous area are continuously detected by the image processing algorithm on the basis of pedestrian detection, and the steps are as follows:
(1) for the detected pedestrian area (x, y, w, h), defining a square area of the head (w, w) of the pedestrian as the head of the pedestrian, and converting pixels in the area from RGB space to HSV space;
(2) in the space, counting the proportion of the pixel points falling in a given color interval; when the proportion exceeds a certain threshold value, judging that the pedestrian correctly wears the safety helmet;
(3) the detection algorithm based on the danger area is based on pedestrian detection, so that for a given danger area (x0, y0, w0, h0) and detected pedestrian area (x, y, w, h), the intersection of the two is calculated according to the plane geometry principle, and the formula is as follows:
NOT(x0≤x+w AND x≤x0+w0AND y0≤y+h AND y≤y0+h0)；
(4) and (4) detecting whether the pedestrian enters a safety area according to the calculation result of the step (3) and carrying out subsequent processing according to the result.
As one of the preferable modes of the invention, the flight control module realizes the control of the flight attitude of the unmanned aerial vehicle by combining a fuzzy PID control algorithm based on system error acceleration compensation and sliding mode control, thereby ensuring the wind resistance and robustness of the system and effectively improving the stability of the flight control system and the tracking precision of the running track;
wherein, the fuzzy PID control algorithm comprises two parts: PID and fuzzy PID controllers; the fuzzy PID controller collects error data and error change conditions, adjusts three parameters of PID in time and acts on a controlled object, the fuzzy PID controller comprises input and output quantity, fuzzy reasoning, an executing mechanism and a control object, and the specific control method comprises the following steps:
(1) determining input variables of the system;
(2) determining the structure of the fuzzy PID controller according to the selected input variable;
(3) fuzzification processing is carried out on the input variable, and the accurate input quantity is converted into a fuzzy quantity;
(4) converting the fuzzy quantity into the control quantity of a control object through fuzzy reasoning and defuzzification according to a set fuzzy logic rule;
(5) the method comprises the steps that self-supervision PID control is optimized and integrated into a self-supervision PID control model to form high-precision PID improved model control, and the automatic supervision control of the unmanned aerial vehicle running track is realized;
the parameter matrix of the unmanned aerial vehicle is represented by the following formula:
wherein M (q) is,
and G0(q) represents a nominal model parameter matrix,. DELTA.M (q),
And Δ g (q) represents the model error, showing the uncertainty of the model parameters, so the model of the unmanned aerial vehicle is:
in the above formula, the first and second carbon atoms are,
representing random interference and errors of the system in the modeling process; suppose qd∈RnWhen the load requirement n of the bounded track is 3, the speed and the acceleration are respectively expressed as the bounded track
And
the following is its tracking definition:
if the random error and the interference influencing the modeling are known, an ideal control law is designed by using a feedback-based linear method, and the expression is as follows:
in formula (4) < lambda >1And (3) representing coefficients of a group of Hurwitz polynomials, and substituting the control law expression (4) into the unmanned aerial vehicle model expression (2) to obtain:
in the formula (5), when t is close to infinity,
infinitely close to 0, q at this timedCan be seen as being similar to q;
therefore, the system error e, the error change rate ec and the error acceleration d (ec) of the variable are intuitively obtained through the supervision adaptive PID model of the formula (5), and the acceleration regulation factor expression is as follows:
in the formula (6), de (k) represents the error change rate at time t, and dde (k) represents the error acceleration at time t;
the sliding mode control is a variable structure control, a change-over switch is adopted to switch over the structures on two sides of a model switching surface, so that the system moves according to a specified state track, and the specific formula is as follows:
assuming some non-linear system:
in the above formula, x is belonged to RnU belongs to R, t belongs to R, a sliding mode switching function s (x) is selected, and when the input quantity is switched according to the relation of a formula (7);
the formula (7) satisfies the following condition:
(1) the sliding mode arrival condition is met: the derivative of the sliding mode switching function is required to be converged to zero, namely the phase track of the sliding mode switching reaches the sliding mode plane within limited time;
(2) when the switching plane has a sliding mode area, i.e. s (x) is 0;
(3) the system has better motion dynamic performance on the sliding mode surface and meets the Lyapunov stability condition.
As one of the preferable modes of the invention, when the server unit is connected with the unmanned aerial vehicle and operates in real time, the default setting is that the mobile terminal cannot be connected with the unmanned aerial vehicle (4G wireless transmission module); when the mobile terminal is connected with the unmanned aerial vehicle (4G wireless transmission module), the default setting server unit cannot be connected with the unmanned aerial vehicle unit
Compared with the prior art, the invention has the advantages that: (1) according to the visual field safety supervision and management system based on the unmanned aerial vehicle, the unmanned aerial vehicle unit can be controlled only by an authorizer through the mobile terminal or the server unit, the operation is easy, and the automation degree is high; (2) the invention has the capability of easy modification and easy expansion; especially, the original network control module of the unmanned aerial vehicle is upgraded into a 4G network control module in the system, so that the stability and the control performance of the system are ensured, the flight of the unmanned aerial vehicle can be remotely controlled, and the flight radius of the unmanned aerial vehicle is enlarged, so that the condition that a manager can control the flight state of the unmanned aerial vehicle and the cloud deck state of a camera at any time and any place through a mobile terminal is ensured; the flight control module ensures the wind resistance and robustness of the flight control system based on the combination of a fuzzy PID control algorithm based on system error acceleration compensation and a sliding mode control algorithm, and effectively improves the stability of the flight control system and the tracking precision of a running track; (3) according to the invention, a 5.8G analog image transmission module of the unmanned aerial vehicle is transformed and upgraded into a 4G network data transmission module, so that the video can be transmitted smoothly in real time; the 4G network is combined with a data acquisition module and a video acquisition module of the unmanned aerial vehicle, so that flight state data, equipment parameters (working states of an engine, an airborne power system and task equipment) and a video signal of a carried camera of the unmanned aerial vehicle are transmitted to a system platform in real time, and managers can master flight states and remote site videos in real time; (4) the invention can remotely control the holder device of the camera through the 4G network, so that the camera can carry out operations such as the zoom-in and zoom-out of the lens, the up-down and left-right rotation operation of the camera, the zoom of the lens and the like, thereby meeting the requirements of various monitoring management of a construction site; meanwhile, video capture and video recording can be carried out on key operation positions or violation sites through an unmanned aerial vehicle, detection of basic target units (pedestrians) in a scene is carried out through a deep learning-based target detection algorithm (YOLO), and then violation phenomena are judged and fed back; (5) the invention realizes the remote operation of the unmanned aerial vehicle, the remote control of the video cradle head, the monitoring of electric quantity, the fixed-point charging, the real-time returning of flight data and image data, the violation identification and the bidirectional alarm of abnormal conditions.
Drawings
FIG. 1 is a system block diagram of the present invention;
FIG. 2 is a block diagram of the fuzzy PID controller architecture of the present invention.
Detailed Description
The following examples are given for the detailed implementation and specific operation of the present invention, but the scope of the present invention is not limited to the following examples.
As shown in fig. 1: a visual on-site safety supervision management system based on unmanned aerial vehicle, the system includes:
an unmanned aerial vehicle unit: the system is used for collecting data information in real time and supervising the site safety; the system comprises an unmanned aerial vehicle power output module, a flight control module, a camera holder unit, a flight data acquisition module, a data storage module, a 4G wireless transmission module and a 4G network control module; the flight data acquisition module is connected with the data storage module and is used for acquiring and storing flight state data and unmanned aerial vehicle equipment parameters; the 4G network control module is connected with the holder control module and the flight control module, so that the flight of the remote control unmanned aerial vehicle is guaranteed, the flight radius of the unmanned aerial vehicle is enlarged, and management personnel can be guaranteed to control the flight state of the unmanned aerial vehicle and the camera holder state at any time and any place through the mobile terminal; the data storage module is connected with the flight data acquisition module and the video acquisition module and is used for storing data information acquired by each module in real time; the 4G wireless transmission module is connected with the data storage module and the server-side 4G wireless transmission module and transmits data back in real time; the unmanned aerial vehicle power output module is connected with each module and supplies power to each module;
a server unit: the system is used for storing data acquired by the unmanned aerial vehicle unit in real time and making corresponding control instruction operation in time; the system comprises a 4G wireless transmission module, a data storage module, a streaming media service module, a flight control service module, an intelligent analysis service module and a server power supply output module; the server power output module is connected with other modules and supplies power to the modules; the streaming media service module is connected with the data storage module and receives real-time video data pushed by the unmanned aerial vehicle in real time; the intelligent analysis service module is connected with the streaming media service module and the data storage module, captures violation information of a site in real time and commands site personnel to modify the violation information; the flight control service module is connected with the data storage module, receives flight data returned by the unmanned aerial vehicle and stores data information; the 4G wireless transmission module is connected with the data storage module and the 4G wireless transmission module at the unmanned aerial vehicle end, so that real-time data communication and storage are realized;
a mobile terminal: the data processing system is used for assisting in storing data returned by the unmanned aerial vehicle, analyzing the data and performing real-time operation, and is connected with the server unit to realize synchronous data sharing.
In order to better supervise the on-site safety through the unmanned aerial vehicle, as one of the preferable modes of the invention, the unmanned aerial vehicle unit further comprises a voice broadcasting module, and the voice broadcasting module is connected with the 4G network control module, and is used for broadcasting background calling content and sending an alarm notice.
In order to improve the safety of the unmanned aerial vehicle, as one of the preferable modes of the unmanned aerial vehicle, the unmanned aerial vehicle unit further comprises a flight electricity testing module, and the flight electricity testing module is connected with the data storage module, so that high-voltage live cables or objects can be automatically avoided, and related constructors can be timely informed of being far away from a dangerous source.
As one of the preferable modes of the invention, the camera pan-tilt unit comprises a video acquisition module and a pan-tilt control module; the video acquisition module is connected with the data storage module and is used for live broadcasting and video recording; and the holder control module is connected with the 4G network control module and is used for controlling the angle of the camera.
As one of the preferable modes of the invention, the mode of the intelligent analysis service module capturing the field violation information in real time is specifically that the basic target unit in the scene is detected through a target detection algorithm based on deep learning; the detection algorithm comprises the following steps:
(1) dividing the image into S-S grids, and when the center of one target falls into a certain grid, the grid is responsible for detecting the corresponding target;
(2) predicting B bounding boxes in each grid, and confidence scores of the bounding boxes, wherein when predicting each bounding box by using the YOLO model, the confidence value of the bounding box is predicted, which can be defined as
The value indicates the reliability of the bounding box containing the object and the reliability of the bounding box;
(3) when the model is trained, if the corresponding grid does not contain the target, the confidence value is equal to 0, otherwise, the confidence value is equal to the predicted box and the IOU of the ground route, and each bounding box contains 5 values: x, y, w, h, confidence; wherein, x and y represent the center of the bounding box, w and h represent the width and height of the bounding box, and confidence represents the IOU of the bounding box and the ground route; the confidence value is predicted and not actually calculated, and each grid predicts C conditional probabilities Pr (Class)i|Object)；
(4) During testing, multiplying the confidence coefficient of the bounding box with the class probability to obtain a confidence score of a specific class; the confidence score represents the probability that the class appears in the bounding box and also represents the degree to which the bounding box matches the target; the calculation formula is as follows:
As one of the preferable modes of the invention, in the mode that the intelligent analysis service module captures the site violation information in real time, after the basic target unit in the scene is detected by the target detection algorithm based on deep learning, the safety helmet and the dangerous area are continuously detected by the image processing algorithm on the basis of pedestrian detection, and the steps are as follows:
(1) for the detected pedestrian area (x, y, w, h), defining a square area of the head (w, w) of the pedestrian as the head of the pedestrian, and converting pixels in the area from RGB space to HSV space;
(2) in the space, counting the proportion of the pixel points falling in a given color interval; when the proportion exceeds a certain threshold value, judging that the pedestrian correctly wears the safety helmet;
(3) the detection algorithm based on the danger area is based on pedestrian detection, so that for a given danger area (x0, y0, w0, h0) and detected pedestrian area (x, y, w, h), the intersection of the two is calculated according to the plane geometry principle, and the formula is as follows:
NOT(x0≤x+w AND x≤x0+w0AND y0≤y+h AND y≤y0+h0)；
(4) and (4) detecting whether the pedestrian enters a safety area according to the calculation result of the step (3) and carrying out subsequent processing according to the result.
In order to improve the stability and robustness of the system, as one of the preferable modes of the invention, the flight control module realizes the control of the flight attitude of the unmanned aerial vehicle in a mode of combining a fuzzy PID control algorithm based on system error acceleration compensation and sliding mode control, thereby ensuring the wind resistance and robustness of the system and effectively improving the stability and the tracking precision of the operation track of the flight control system;
wherein, the fuzzy PID control algorithm comprises two parts: PID and fuzzy PID controllers; the fuzzy PID controller collects error data and error change conditions, adjusts three parameters of PID in time, and acts on a controlled object, as shown in FIG. 2: the fuzzy PID controller comprises input and output quantity, fuzzy inference, an actuating mechanism and a control object, and the specific control method comprises the following steps:
(1) determining input variables of the system;
(2) determining the structure of the fuzzy PID controller according to the selected input variable;
(3) fuzzification processing is carried out on the input variable, and the accurate input quantity is converted into a fuzzy quantity;
(4) converting the fuzzy quantity into the control quantity of a control object through fuzzy reasoning and defuzzification according to a set fuzzy logic rule;
(5) the method comprises the steps that self-supervision PID control is optimized and integrated into a self-supervision PID control model to form high-precision PID improved model control, and the automatic supervision control of the unmanned aerial vehicle running track is realized;
the parameter matrix of the unmanned aerial vehicle is represented by the following formula:
wherein M (q) is,
and G0(q) represents a nominal model parameter matrix,. DELTA.M (q),
And Δ g (q) represents the model error, showing the uncertainty of the model parameters, so the model of the unmanned aerial vehicle is:
in the above formula, the first and second carbon atoms are,
representing random interference and errors of the system in the modeling process; suppose qd∈RnWhen the load requirement n of the bounded track is 3, the speed and the acceleration are respectively expressed as the bounded track
And
the following is its tracking definition:
if the random error and the interference influencing the modeling are known, an ideal control law is designed by using a feedback-based linear method, and the expression is as follows:
in formula (4) < lambda >1And (3) representing coefficients of a group of Hurwitz polynomials, and substituting the control law expression (4) into the unmanned aerial vehicle model expression (2) to obtain:
in the formula (5), when t is close to infinity,
infinitely close to 0, at which point qd can be considered to be approximately q;
therefore, the system error e, the error change rate ec and the error acceleration d (ec) of the variable are intuitively obtained through the supervision adaptive PID model of the formula (5), and the acceleration regulation factor expression is as follows:
in the formula (6), de (k) represents the error change rate at time t, and dde (k) represents the error acceleration at time t; (ii) a
The fuzzy compensation method based on error acceleration is adopted to improve the unmanned aerial vehicle supervision adaptive PID model to obtain a high-precision PID improved model, the precision of robot track control is improved, the fuzzy compensation method is similar to the supervision adaptive PID structure, and the difference is the difference of output variables based on the difference of the output variablesObtaining an acceleration adjusting factor R (k) and an error absolute value | e | by an acceleration fuzzy compensation principle; the output variable of the supervised adaptive PID structure is Rp、RiAnd RdThe fuzzy compensation based on the error acceleration can reflect the system response condition, so that the supervision self-adaptive PID model can be controlled more accurately.
The sliding mode control is called sliding mode variable structure control, and the main idea is that the system starts from a set state, reaches a sliding mode surface within limited time, generates sliding motion on the sliding mode surface, and finally reaches a balance point; the sliding mode control has low dependence degree on a specific model and strong anti-interference capability, and is widely applied to the control of a nonlinear system.
The sliding mode control is a variable structure control, a change-over switch is adopted to switch over the structures on two sides of a model switching surface, so that the system moves according to a specified state track, and the specific formula is as follows:
assuming some non-linear system:
in the above formula, x is belonged to RnU belongs to R, t belongs to R, a sliding mode switching function s (x) is selected, and when the input quantity is switched according to the relation of a formula (7);
the formula (7) satisfies the following condition:
(1) the sliding mode arrival condition is met: the derivative of the sliding mode switching function is required to be converged to zero, namely the phase track of the sliding mode switching reaches the sliding mode plane within limited time;
(2) when the switching plane has a sliding mode area, i.e. s (x) is 0;
(3) the system has better motion dynamic performance on the sliding mode surface and meets the Lyapunov stability condition.
Through the combination of the two algorithms, the wind resistance and the robustness of the system are ensured, and the stability of the flight control system and the tracking precision of the operation track are effectively improved.
In order to maintain the stability of the system, as one of the preferable modes of the invention, when the server unit is connected with the unmanned aerial vehicle and operates in real time, the default setting is that the mobile terminal cannot be connected with the unmanned aerial vehicle (4G wireless transmission module); when the mobile terminal is connected with the unmanned aerial vehicle (4G wireless transmission module), the default setting server unit cannot be connected with the unmanned aerial vehicle unit
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (7)
1. The utility model provides a visual scene safety supervision management system based on unmanned aerial vehicle which characterized in that, this system includes:
an unmanned aerial vehicle unit: the system is used for collecting data information in real time and supervising the site safety;
a server unit: the system is used for storing data acquired by the unmanned aerial vehicle unit in real time and making corresponding control instruction operation in time;
a mobile terminal: the system is used for assisting in storing data returned by the unmanned aerial vehicle, analyzing the data and performing real-time operation, and is connected with the server unit to realize synchronous sharing of the data;
the server unit includes: the system comprises a 4G wireless transmission module, a data storage module, a streaming media service module, a flight control service module, an intelligent analysis service module and a server power supply output module; the server power output module is connected with other modules and supplies power to the modules; the streaming media service module is connected with the data storage module and receives real-time video data pushed by the unmanned aerial vehicle in real time; the intelligent analysis service module is connected with the streaming media service module and the data storage module, captures violation information of a site in real time and commands site personnel to modify the violation information; the flight control service module is connected with the data storage module, receives flight data returned by the unmanned aerial vehicle and stores data information; the 4G wireless transmission module is connected with the data storage module and the 4G wireless transmission module at the unmanned aerial vehicle end, so that real-time data communication and storage are realized;
the intelligent analysis service module captures the on-site violation information in real time in a mode of detecting a basic target unit in a scene through a target detection algorithm based on deep learning; the detection algorithm comprises the following steps:
(1) dividing the image into S-S grids, and when the center of one target falls into a certain grid, the grid is responsible for detecting the corresponding target;
(2) predicting B bounding boxes in each grid, and confidence scores of the bounding boxes, wherein when predicting each bounding box by using the YOLO model, the confidence value of the bounding box is predicted, which can be defined as
The value indicates the reliability of the bounding box containing the object and the reliability of the bounding box;
(3) when the model is trained, if the corresponding grid does not contain the target, the confidence value is equal to 0, otherwise, the confidence value is equal to the predicted box and the IOU of the ground route, and each bounding box contains 5 values: x, y, w, h, confidence; wherein, x and y represent the center of the bounding box, w and h represent the width and height of the bounding box, and confidence represents the IOU of the bounding box and the ground route; the confidence value is predicted and not actually calculated, and each grid predicts C conditional probabilities Pr (Class)i|Object)；
(4) During testing, multiplying the confidence coefficient of the bounding box with the class probability to obtain a confidence score of a specific class; the confidence score represents the probability that the class appears in the bounding box and also represents the degree to which the bounding box matches the target; the calculation formula is as follows:
in the mode that the intelligent analysis service module captures the field violation information in real time, after the basic target unit in the scene is detected through a target detection algorithm based on deep learning, on the basis of pedestrian detection, the safety helmet and the dangerous area are continuously detected through an image processing algorithm, and the method comprises the following steps:
(1) for the detected pedestrian area (x, y, w, h), defining a square area of the head (w, w) of the pedestrian as the head of the pedestrian, and converting pixels in the area from RGB space to HSV space;
(2) in the space, counting the proportion of the pixel points falling in a given color interval; when the proportion exceeds a certain threshold value, judging that the pedestrian correctly wears the safety helmet;
(3) the detection algorithm based on the danger area is based on pedestrian detection, so that for a given danger area (x0, y0, w0, h0) and detected pedestrian area (x, y, w, h), the intersection of the two is calculated according to the plane geometry principle, and the formula is as follows:
NOT(x0≤x+w AND x≤x0+w0 ANDy0≤y+h ANDy≤y0+h0)；
(4) and (4) detecting whether the pedestrian enters a safety area according to the calculation result of the step (3) and carrying out subsequent processing according to the result.
2. The unmanned aerial vehicle-based visual field security surveillance management system of claim 1, wherein the unmanned aerial vehicle unit comprises: the system comprises an unmanned aerial vehicle power output module, a flight control module, a camera holder unit, a flight data acquisition module, a data storage module, a 4G wireless transmission module and a 4G network control module; the camera pan-tilt unit comprises a video acquisition module and a pan-tilt control module; the flight data acquisition module is connected with the data storage module and is used for acquiring and storing flight state data and unmanned aerial vehicle equipment parameters; the 4G network control module is connected with the holder control module and the flight control module; the data storage module is connected with the flight data acquisition module and the video acquisition module and is used for storing data information acquired by each module in real time; the 4G wireless transmission module is connected with the data storage module and the server-side 4G wireless transmission module and transmits data back in real time; the unmanned aerial vehicle power output module is connected with each module and supplies power for each module.
3. The visual field safety supervision and management system based on unmanned aerial vehicle of claim 2, characterized in that the unmanned aerial vehicle unit further comprises a voice broadcasting module, and the voice broadcasting module is connected with the 4G network control module and is used for broadcasting background shouting content and sending out alarm notification.
4. The unmanned aerial vehicle-based visual field safety supervision management system according to claim 2, wherein the unmanned aerial vehicle unit further comprises a flight electricity testing module, and the flight electricity testing module is connected with the data storage module and is used for automatically avoiding high-voltage live cables or objects and timely informing relevant constructors of keeping away from danger sources.
5. The visual field safety supervision management system based on unmanned aerial vehicle of claim 2, characterized in that the video acquisition module is connected with the data storage module for live broadcast and video recording; and the holder control module is connected with the 4G network control module and is used for controlling the angle of the camera.
6. The visual field safety supervision management system based on the unmanned aerial vehicle according to claim 1, characterized in that the flight control module realizes the control of the flight attitude of the unmanned aerial vehicle by combining a fuzzy PID control algorithm based on system error acceleration compensation and sliding mode control;
wherein, the fuzzy PID control algorithm comprises two parts: PID and fuzzy PID controllers; the fuzzy PID controller collects error data and error change conditions, adjusts three parameters of PID in time and acts on a controlled object, the fuzzy PID controller comprises input and output quantity, fuzzy reasoning, an executing mechanism and a control object, and the specific control method comprises the following steps:
(1) determining input variables of the system;
(2) determining the structure of the fuzzy PID controller according to the selected input variable;
(3) fuzzification processing is carried out on the input variable, and the accurate input variable is converted into a fuzzy quantity;
(4) converting the fuzzy quantity into the control quantity of a control object through fuzzy reasoning and defuzzification according to a set fuzzy logic rule;
(5) the method comprises the steps that self-supervision PID control is optimized and integrated into a self-supervision PID control model to form high-precision PID improved model control, and the automatic supervision control of the unmanned aerial vehicle running track is realized;
the parameter matrix of the unmanned aerial vehicle is represented by the following formula:
wherein M (q) is,
and G0(q) represents a nominal model parameter matrix,. DELTA.M (q),
And Δ g (q) represents the model error, showing the uncertainty of the model parameters, so the model of the unmanned aerial vehicle is:
in the above formula, the first and second carbon atoms are,
representing random interference and errors of the system in the modeling process; suppose qd∈RnWhen the load requirement n of the bounded track is 3, the speed and the acceleration are respectively expressed as the bounded track
The following is its tracking definition:
if the random error and the interference influencing the modeling are known, an ideal control law is designed by using a feedback-based linear method, and the expression is as follows:
in formula (4) < lambda >1And (3) representing coefficients of a group of Hurwitz polynomials, and substituting the control law expression (4) into the unmanned aerial vehicle model expression (2) to obtain:
in the formula (5), when t is close to infinity,
infinitely close to 0, q at this timedCan be seen as being similar to q;
therefore, the system error e, the error change rate ec and the error acceleration dec (k) of the variables are intuitively obtained through the supervision adaptive PID model of the formula (5), and the acceleration regulating factor is expressed as follows:
in the formula (6), de (k) represents the error change rate at time t, and dde (k) represents the error acceleration at time t;
the sliding mode control is a variable structure control, a change-over switch is adopted to switch over the structures on two sides of a model switching surface, so that the system moves according to a specified state track, and the specific formula is as follows:
assuming some non-linear system:
in the above formula, x is belonged to RnU belongs to R, t belongs to R, a sliding mode switching function s (x) is selected, and when the input quantity is switched according to the relation of a formula (7);
the formula (7) satisfies the following condition:
(1) the sliding mode arrival condition is met: the derivative of the sliding mode switching function is required to be converged to zero, namely the phase track of the sliding mode switching reaches the sliding mode plane within limited time;
(2) when the switching plane has a sliding mode area, i.e. s (x) is 0;
(3) the system has better motion dynamic performance on the sliding mode surface and meets the Lyapunov stability condition.
7. The visual scene safety supervision management system based on unmanned aerial vehicle of any one of claims 1-6, characterized in that when the server unit is connected with unmanned aerial vehicle and operates in real time, the default setting mobile terminal can not be connected with unmanned aerial vehicle; when the mobile terminal is connected with the unmanned aerial vehicle, the default setting server unit cannot be connected with the unmanned aerial vehicle unit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810467918.5A CN108445767B (en) | 2018-05-16 | 2018-05-16 | Visual on-site safety supervision and management system based on unmanned aerial vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810467918.5A CN108445767B (en) | 2018-05-16 | 2018-05-16 | Visual on-site safety supervision and management system based on unmanned aerial vehicle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108445767A CN108445767A (en) | 2018-08-24 |
| CN108445767B true CN108445767B (en) | 2021-04-27 |
Family
ID=63204229
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810467918.5A Active CN108445767B (en) | 2018-05-16 | 2018-05-16 | Visual on-site safety supervision and management system based on unmanned aerial vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108445767B (en) |
Families Citing this family (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109559006A (en) * | 2018-09-19 | 2019-04-02 | 中建科技有限公司深圳分公司 | A kind of monitoring method, device and the terminal device of the construction site based on unmanned plane |
| CN109558783A (en) * | 2018-09-20 | 2019-04-02 | 中建科技有限公司深圳分公司 | A kind of violation detection method, system and equipment for construction site |
| WO2020056677A1 (en) * | 2018-09-20 | 2020-03-26 | 中建科技有限公司深圳分公司 | Violation detection method, system, and device for building construction site |
| CN109358654B (en) * | 2018-11-16 | 2022-03-29 | 江苏科技大学 | Water-air amphibious search and rescue support unmanned aerial vehicle system |
| CN109712453A (en) * | 2019-02-28 | 2019-05-03 | 安徽腾策网络科技有限公司 | A kind of software Training Management Information System based on big data |
| CN109870910A (en) * | 2019-03-02 | 2019-06-11 | 哈尔滨理工大学 | A kind of flying vehicles control method based on synovial membrane control |
| CN110147114A (en) * | 2019-05-30 | 2019-08-20 | 国网福建省电力有限公司莆田供电公司 | A kind of scene based on unmanned plane auxiliary monitoring system and method violating the regulations |
| CN111741254B (en) * | 2020-05-13 | 2022-03-25 | 苏州锦添科技产业发展有限公司 | Visual digital twin high-end equipment system based on unmanned aerial vehicle information terminal |
| CN111614969A (en) * | 2020-05-14 | 2020-09-01 | 深圳供电局有限公司 | Unmanned aerial vehicle tour video live broadcast method and system |
| CN112214029A (en) * | 2020-09-10 | 2021-01-12 | 江苏久飞智能科技有限公司 | Airborne SOA type task management calculation control system of power inspection unmanned aerial vehicle |
| CN112700225A (en) * | 2021-01-07 | 2021-04-23 | 湖南联智科技股份有限公司 | Technical architecture for unmanned aerial vehicle cruise monitoring platform management |
| CN112804547B (en) * | 2021-01-07 | 2022-08-23 | 河北交通职业技术学院 | Interactive live broadcast system based on unmanned aerial vehicle VR makes a video recording |
| CN115454138B (en) * | 2022-10-11 | 2023-04-18 | 众芯汉创（北京）科技有限公司 | Construction violation determination method and system based on unmanned aerial vehicle image recognition technology |
| CN116012787A (en) * | 2023-01-10 | 2023-04-25 | 山东高速建设管理集团有限公司 | Safety monitoring method and system based on high-altitude balloon and unmanned aerial vehicle bee colony |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202632112U (en) * | 2012-06-29 | 2012-12-26 | 山东电力集团公司电力科学研究院 | Electric field difference obstacle avoidance system for live wire tour inspection of unmanned aerial vehicle |
| CN103135550A (en) * | 2013-01-31 | 2013-06-05 | 南京航空航天大学 | Multiple obstacle-avoidance control method of unmanned plane used for electric wire inspection |
| CN104898596A (en) * | 2015-04-02 | 2015-09-09 | 中国石油大学(华东) | On-site multifunctional safety supervision integrated management system |
| CN106791712A (en) * | 2017-02-16 | 2017-05-31 | 周欣 | A kind of monitoring system and method in construction site |
| CN206322023U (en) * | 2016-11-25 | 2017-07-11 | 武汉理工大学 | A kind of unmanned plane villa community is violating the regulations to build supervising device |
| CN107274510A (en) * | 2017-06-08 | 2017-10-20 | 广东容祺智能科技有限公司 | A kind of unmanned plane Power Line Inspection System based on electromagnetism avoidance |
| CN107450589A (en) * | 2017-07-25 | 2017-12-08 | 广东容祺智能科技有限公司 | A kind of construction safety based on unmanned plane is maked an inspection tour and early warning system |
| CN107734296A (en) * | 2017-09-30 | 2018-02-23 | 贵州电网有限责任公司铜仁供电局 | A kind of capital construction scene monitoring unmanned system |
-
2018
- 2018-05-16 CN CN201810467918.5A patent/CN108445767B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202632112U (en) * | 2012-06-29 | 2012-12-26 | 山东电力集团公司电力科学研究院 | Electric field difference obstacle avoidance system for live wire tour inspection of unmanned aerial vehicle |
| CN103135550A (en) * | 2013-01-31 | 2013-06-05 | 南京航空航天大学 | Multiple obstacle-avoidance control method of unmanned plane used for electric wire inspection |
| CN104898596A (en) * | 2015-04-02 | 2015-09-09 | 中国石油大学(华东) | On-site multifunctional safety supervision integrated management system |
| CN206322023U (en) * | 2016-11-25 | 2017-07-11 | 武汉理工大学 | A kind of unmanned plane villa community is violating the regulations to build supervising device |
| CN106791712A (en) * | 2017-02-16 | 2017-05-31 | 周欣 | A kind of monitoring system and method in construction site |
| CN107274510A (en) * | 2017-06-08 | 2017-10-20 | 广东容祺智能科技有限公司 | A kind of unmanned plane Power Line Inspection System based on electromagnetism avoidance |
| CN107450589A (en) * | 2017-07-25 | 2017-12-08 | 广东容祺智能科技有限公司 | A kind of construction safety based on unmanned plane is maked an inspection tour and early warning system |
| CN107734296A (en) * | 2017-09-30 | 2018-02-23 | 贵州电网有限责任公司铜仁供电局 | A kind of capital construction scene monitoring unmanned system |
Non-Patent Citations (2)
| Title |
|---|
| 无人机在建筑施工现场的应用研究;李彩霞;《中国优秀硕士学位论文全文数据库（电子期刊）》;20170215(第02期);第C038-2251页 * |
| 李彩霞.无人机在建筑施工现场的应用研究.《中国优秀硕士学位论文全文数据库（电子期刊）》.2017,(第02期),第C038-2251页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108445767A (en) | 2018-08-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108445767B (en) | Visual on-site safety supervision and management system based on unmanned aerial vehicle | |
| CN110488841B (en) | Transformer equipment combined inspection system based on intelligent robot and application method thereof | |
| CN112910094B (en) | Remote automatic transformer substation inspection system and method based on ubiquitous power Internet of things | |
| CN108189043B (en) | Inspection method and inspection robot system applied to high-speed rail machine room | |
| CN109284739B (en) | Power transmission line external damage prevention early warning method and system based on deep learning | |
| US11561251B2 (en) | Remote autonomous inspection of utility system components utilizing drones and rovers | |
| US7109875B2 (en) | Sensor network system managing method, sensor network system managing program, storage medium containing sensor network system managing program, sensor network system managing device, relay network managing method, relay network managing program, storage medium containing relay network managing program, and relay network managing device | |
| CN109861387B (en) | Intelligent inspection system for transformer substation | |
| CN116755474A (en) | Electric power line inspection method and system for unmanned aerial vehicle | |
| CN114250471A (en) | Cathodic protection potential follow-up control system under oil gas wisdom pipeline framework | |
| CN113569914A (en) | Power transmission line inspection method and system fusing point cloud data | |
| CN114115020A (en) | Intelligent control system and control method for height of unmanned aerial vehicle | |
| CN112332541A (en) | Monitoring system and method for transformer substation | |
| CN115271113A (en) | Intelligent inspection method for construction process of pumped storage power station | |
| CN116360440A (en) | Unmanned ship ground station interaction system | |
| CN114510077A (en) | Route planning method and device for unmanned aerial vehicle pole routing inspection and computer storage medium | |
| CN105811578A (en) | Power transmission line monitoring platform, power source monitoring algorithm thereof and image early warning algorithm of power transmission line monitoring platform | |
| CN117522028A (en) | Visual display system for real-time operation data of hydropower station | |
| CN117111660A (en) | Unattended intelligent granary system and method | |
| CN117030032A (en) | Equipment part temperature measurement method and device, electronic equipment and storage medium | |
| CN115457411B (en) | Unmanned inspection method and device for aviation oil pipeline and aviation oil safety operation and maintenance system | |
| CN205911826U (en) | Open digital integrated automation system for transformer substation | |
| CN113541314A (en) | Transformer substation combined inspection system and control method thereof | |
| CN110430437B (en) | Self-adaptive cloud storage method of remote vision system | |
| CN114202910A (en) | Instrument recognition device, instrument monitoring system and monitoring method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |