Aircraft high-precision positioning method and device based on opportunity signal fusion

Abstract

The present invention relates to the field of high-precision positioning of aircraft, and specifically to a high-precision positioning method and device for aircraft based on opportunity signal fusion, comprising: when the aircraft signal strength is lower than a preset threshold, collecting opportunity signals through broadcast, Wi-Fi, radar, and Bluetooth signal receivers; extracting multi-dimensional signal features and outputting a signal source quality score using a deep Q-network reinforcement learning model; treating signal weights as state variables and dynamically adjusting weights in combination with a Kalman filter algorithm; and using triangulation and multilateral positioning to determine the aircraft's position coordinates and weightedly calculate the final position based on the signal source weights and distance information. The device includes opportunity signal collection, signal feature extraction, dynamic weight adjustment, and positioning settlement modules, each of which performs corresponding functions; the present invention integrates multiple opportunity signals to improve the positioning accuracy and reliability of aircraft in complex environments.

Description

Technical Field

[0001] The present invention relates to the field of high-precision positioning of aircraft, and in particular to a method and device for high-precision positioning of aircraft based on opportunity signal fusion. Background Art

[0002] During the operation of an aircraft, the accuracy of positioning plays a decisive role in its safe flight and mission execution. Traditional aircraft positioning mainly relies on GPS, which can provide the aircraft with relatively accurate positioning data in an environment with good signals to meet general flight needs.

[0003] However, in actual flight scenarios, aircraft often face various complex environments. In urban areas, tall buildings block and refract signals, causing multipath effects that seriously interfere with normal signal reception and significantly reduce positioning accuracy. In areas with complex electromagnetic environments, such as areas with dense communication base stations or strong electromagnetic interference sources, signals are interfered with, making positioning accuracy impossible to guarantee.

[0004] Traditional single-based positioning methods can no longer meet the requirements of these complex tasks. In order to solve the above defects, a technical solution is now provided. Summary of the Invention

[0005] In order to solve the technical problems raised by the above background technology, the present invention provides a method and device for high-precision positioning of an aircraft based on opportunity signal fusion.

[0006] The object of the present invention can be achieved through the following technical solutions: In a first aspect, the present invention provides a method for high-precision positioning of an aircraft based on opportunity signal fusion, the specific steps of which are as follows:

[0007] Step 1: Opportunity signal acquisition: When the aircraft is performing a flight mission, the monitoring program obtains the signal strength of the current aircraft in real time. When the signal strength is lower than the preset signal strength threshold, a signal acquisition instruction is generated and distributed to multiple signal receiving devices. The multiple signal receiving devices include broadcast signal receivers, Wi-Fi signal receivers, radar signal receivers, and Bluetooth signal receivers. The multiple signal receiving devices of the aircraft obtain the signal acquisition instruction and collect various opportunity signals in the surrounding environment from the broadcast signal receivers, Wi-Fi signal receivers, radar signal receivers, and Bluetooth signal receivers in real time. The signals are then transmitted to the aircraft through the tuning circuit of the broadcast signal receiver. The system scans the medium wave and short wave frequency bands. When a broadcast signal is detected, the front-end circuit of the receiver amplifies and filters the broadcast signal to remove noise interference and obtain a processed broadcast signal. The wireless frequency band is then scanned through the Wi-Fi signal receiver. When a hotspot signal is detected, the MAC address of the current hotspot is first identified, the signal strength is detected through the RSSI circuit, and multiple samples are collected from the same hotspot at preset intervals to obtain the signal strength values and calculate the average signal strength. If the signal strength average is greater than the pre-set signal strength threshold, it is determined to be the target signal and sent to the signal feature extraction. Otherwise, the signal source is deleted.

[0008] By analogy, a radar transmitter is set next to the radar signal receiver, and electromagnetic wave pulses are sent through the radar transmitter. The radar signal receiver receives and amplifies the transmitted echo signal; finally, the Bluetooth signal receiver scans multiple Bluetooth signals according to the preset interval, and obtains the signal strength of each Bluetooth device, arranges the signal strengths, obtains the Bluetooth device with the largest signal strength and marks it as the first device.

[0009] Step 2: Multi-dimensional signal feature extraction: Extract the features of the broadcast signal receiver, Wi-Fi signal receiver, radar signal receiver, and Bluetooth signal receiver, track the carrier phase in the broadcast signal based on PLL, and use the Allan variance formula Get the carrier phase stability value ,in, is the carrier frequency, M is the number of measurements, It is The average frequency within a measurement interval; the RF signal power is converted into a voltage value through a logarithmic amplifier, and then sampled by an analog-to-digital converter (ADC) to calculate the broadcast signal strength sp; then the current signal is demodulated to obtain a real-time modulation signal , extract the standard modulation signal in the database , using the mean square error formula Get the modulation distortion value MSE;

[0010] Use a spectrum analyzer to detect the interference signal power in the WiFi target channel, and then obtain the target signal power. Divide the interference signal power by the target signal power to obtain the co-channel interference value.

[0011] By analogy, the radar signal is feature extracted and the time difference between the radar transmitting pulse and the echo pulse is marked as , calculated according to the formula Get the distance d1 between the aircraft and the target radar, where c represents the speed of light; then use the formula to calculate Get the Doppler frequency shift generated by the echo signal , where v represents the relative speed between the aircraft and the radar, Indicates the signal wavelength, Indicates the relative motion angle between the radar and the aircraft;

[0012] Finally, the Bluetooth signal features are extracted, the start time of the signal acquisition of the first device is marked, and the end time of the acquisition is marked. The time difference between the two marked time points is calculated to obtain the acquisition period, and the number of signal acquisitions within the acquisition period and the corresponding signal value are marked as ,in , j represents the signal acquisition number, m represents the total number of acquisitions, and the mean value RE of the Bluetooth signal is calculated by the mean value, and then the standard deviation formula is calculated Get the Bluetooth signal standard deviation RQ. Low variance indicates a stable signal, which usually corresponds to direct line-of-sight path propagation.

[0013] Furthermore, a reinforcement learning model is established using a deep Q network structure, which includes an input layer, multiple hidden layers and an output layer. The input layer receives the extracted multi-dimensional signal features, the hidden layer performs nonlinear transformation and abstraction on the features, and the output layer outputs the quality score of each signal source. Specifically, the extracted multi-dimensional signal features are combined into a state vector s as the input of the reinforcement learning model, and then the model's actions are set. Different signal sources are assigned different quality scores, with values ranging from 0 to 1. Then, a reward function is set for the reinforcement learning model, and the quality score vector of the current time step is , the quality score vector of the previous time step is , then the rate of change of the quality score can be expressed as , extract the preset signal stability threshold in the database ,like When the signal is less than the preset signal stability threshold, additional rewards will be given , on the contrary, there is no reward, and the reward function is expressed as ; The reinforcement learning model calculates each action through forward propagation based on the input state vector s The Q value represents the action taken in state s The model selects the action with the largest Q value as the output, that is, the quality score value corresponding to each signal source is obtained. It should be noted that the number of signal sources is not fixed. The quality score value of the corresponding number is output based on the number of signal sources currently collected by the aircraft in real time.

[0014] Step 3: Dynamic weight adjustment: Consider the weight of each signal as a state variable, construct the observation equation space, and then combine the weight adjustment of the Kalman filter algorithm. Specifically, obtain the number of signal sources currently received and mark it as g, and the weight vector is expressed as Z=[z1,z2,……,z g ], set the state transfer equation to , Expressed as the weight change, the observation equation space is constructed accordingly;

[0015] According to the weight estimate of the previous moment And the state transfer equation calculates the prediction weight, the formula is , and the prediction covariance formula is ,in represents the covariance of the previous moment, Represents the process noise covariance, converts each quality score value into a quality assessment vector and integrates it into the quality assessment result , using the Kalman gain formula Get the Kalman gain value at that moment ,in Represents the observation matrix obtained by mapping the weight state to the observation equation space. Similarly Expressed as the observation noise covariance, the weight estimate is then updated and covariance ,Through continuous iterative prediction and updating steps, the weight of each signal source is adjusted and transmitted to the positioning settlement based on the changes of the signal features collected and extracted in real time.

[0016] Step 4: Positioning and Settlement: Based on the corresponding weights of each signal source, the distance information between multiple signal sources and the aircraft is used to determine the position coordinates of the aircraft through geometric positioning principles, including triangulation and multilateral positioning. Specifically, the free space propagation model is set as the distance measurement formula, and the transmission power, signal wavelength, and receiving power of the broadcast signal receiver, Wi-Fi signal receiver, and Bluetooth signal receiver are obtained, marked as 、 and , substitute the broadcast signal receiver's transmission power, signal wavelength, and receiving power into the formula to calculate Get the broadcast signal distance value d2, and use the same steps and formulas to get the Wi-Fi signal distance value d3 and the Bluetooth signal distance value d4;

[0017] With the aircraft as the center point, when three signal sources are obtained, their coordinates are marked as 、 and Calculated based on the triangulation principle Get the two-dimensional position coordinates of the aircraft ; If four signal sources are obtained, their coordinates are marked as 、 、 and , calculated using the principle of multilateration Get the three-dimensional position coordinates of the aircraft ;

[0018] Taking three-dimensional coordinates as an example, the weights Q1, Q2 and Q3 of each signal source obtained by dynamic weight adjustment are as follows: 、 and , calculate the three-dimensional coordinates of the aircraft and the corresponding weights Get the final aircraft position .

[0019] A second aspect of the present invention provides a high-precision positioning device for an aircraft based on opportunity signal fusion, comprising:

[0020] The opportunity signal acquisition module monitors the signal strength of the aircraft in real time during the aircraft's flight mission. When the signal strength falls below a preset threshold, it generates a signal acquisition instruction and distributes it to multiple signal receiving devices, which then collect the signal.

[0021] The signal feature extraction module extracts features from broadcast signal receivers, Wi-Fi signal receivers, radar signal receivers, and Bluetooth signal receivers, inputs the signal features into a preset reinforcement learning model, and outputs a quality score for each signal source;

[0022] The weight dynamic adjustment module regards the weight of each signal as a state variable, constructs the observation equation space, and then adjusts the weight in real time in combination with the Kalman filter algorithm;

[0023] The positioning settlement module uses the distance information between multiple signal sources and the aircraft based on the corresponding weights of each signal source, and then uses geometric positioning principles, including triangulation and multilateral positioning, to determine the position coordinates of the aircraft, and obtains the final position coordinates based on weighted calculation.

[0024] Compared with existing technologies, the present invention offers the following advantages: During signal processing, it extracts multi-dimensional features from collected signals and uses a reinforcement learning model to score each signal source, screening for high-quality and stable signals. During the signal feature extraction phase, it comprehensively assesses signal quality by calculating indicators such as the carrier phase stability and modulation distortion of broadcast signals, as well as the co-channel interference of Wi-Fi signals. This reduces the interference of low-quality signals on positioning results, further improving positioning accuracy and stability.

[0025] Dynamic weight adjustment treats signal weights as state variables and uses a Kalman filter algorithm to adjust weights in real time. This dynamic adjustment mechanism automatically optimizes the weight distribution of each signal source in the positioning calculation based on real-time changes in signal characteristics. When the signal quality of a signal source improves, its weight is increased, and vice versa. This ensures optimal positioning under varying flight conditions and enhances the positioning system's adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. The following drawings are not intentionally scaled to the actual size, and the focus is on illustrating the main purpose of the present invention.

[0027] Figure 1 It is a flowchart of the method steps of the present invention.

[0028] Figure 2 This is a device block diagram of the present invention.

[0029] Figure 3 This is a structural diagram of the reinforcement learning model of the present invention.

[0030] Figure 4 This is a schematic diagram of positioning settlement. DETAILED DESCRIPTION

[0031] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts also fall within the scope of protection of the present invention.

[0032] Please refer to Figure 1 As shown, the first aspect of the present invention provides a method for high-precision positioning of an aircraft based on opportunity signal fusion, the specific steps of which are as follows:

[0033] Step 1: Opportunity signal acquisition: When the aircraft is performing a flight mission, the monitoring program obtains the signal strength of the current aircraft in real time. When the signal strength is lower than the preset signal strength threshold, a signal acquisition instruction is generated and distributed to multiple signal receiving devices. The multiple signal receiving devices include broadcast signal receivers, Wi-Fi signal receivers, radar signal receivers, and Bluetooth signal receivers. The multiple signal receiving devices of the aircraft obtain the signal acquisition instruction and collect various opportunity signals in the surrounding environment from the broadcast signal receivers, Wi-Fi signal receivers, radar signal receivers, and Bluetooth signal receivers in real time. The signals are then transmitted to the aircraft through the tuning circuit of the broadcast signal receiver. The system scans the medium wave and short wave frequency bands. When a broadcast signal is detected, the front-end circuit of the receiver amplifies and filters the broadcast signal to remove noise interference and obtain a processed broadcast signal. The wireless frequency band is then scanned through the Wi-Fi signal receiver. When a hotspot signal is detected, the MAC address of the current hotspot is first identified, the signal strength is detected through the RSSI circuit, and multiple samples are collected from the same hotspot at preset intervals to obtain the signal strength values and calculate the average signal strength. If the signal strength average is greater than the pre-set signal strength threshold, it is determined to be the target signal and sent to the signal feature extraction. Otherwise, the signal source is deleted.

[0034] By analogy, a radar transmitter is set next to the radar signal receiver, and electromagnetic wave pulses are sent through the radar transmitter. The radar signal receiver receives and amplifies the transmitted echo signal; finally, the Bluetooth signal receiver scans multiple Bluetooth signals according to the preset interval, and obtains the signal strength of each Bluetooth device, arranges the signal strengths, obtains the Bluetooth device with the largest signal strength and marks it as the first device.

[0035] Step 2: Multi-dimensional signal feature extraction: Extract the features of the broadcast signal receiver, Wi-Fi signal receiver, radar signal receiver, and Bluetooth signal receiver, track the carrier phase in the broadcast signal based on PLL, and use the Allan variance formula Get the carrier phase stability value ,in, is the carrier frequency, M is the number of measurements, It is The average frequency within a measurement interval; the RF signal power is converted into a voltage value through a logarithmic amplifier, and then sampled by an analog-to-digital converter (ADC) to calculate the broadcast signal strength sp; then the current signal is demodulated to obtain a real-time modulation signal , extract the standard modulation signal in the database , using the mean square error formula Get the modulation distortion value MSE;

[0036] Use a spectrum analyzer to detect the interference signal power in the WiFi target channel, and then obtain the target signal power. Divide the interference signal power by the target signal power to obtain the co-channel interference value.

[0037] By analogy, the radar signal is feature extracted and the time difference between the radar transmitting pulse and the echo pulse is marked as , calculated according to the formula Get the distance d1 between the aircraft and the target radar, where c represents the speed of light; then use the formula to calculate Get the Doppler frequency shift generated by the echo signal , where v represents the relative speed between the aircraft and the radar, Indicates the signal wavelength, Indicates the relative motion angle between the radar and the aircraft;

[0038] Finally, the Bluetooth signal features are extracted, the start time of the signal acquisition of the first device is marked, and the end time of the acquisition is marked. The time difference between the two marked time points is calculated to obtain the acquisition period, and the number of signal acquisitions within the acquisition period and the corresponding signal value are marked as ,in , j represents the signal acquisition number, m represents the total number of acquisitions, and the mean value RE of the Bluetooth signal is calculated by the mean value, and then the standard deviation formula is calculated Get the Bluetooth signal standard deviation RQ. Low variance indicates a stable signal, which usually corresponds to direct line-of-sight path propagation.

[0039] Furthermore, a reinforcement learning model is established using a deep Q network structure, which includes an input layer, multiple hidden layers and an output layer. The input layer receives the extracted multi-dimensional signal features, the hidden layer performs nonlinear transformation and abstraction on the features, and the output layer outputs the quality score of each signal source. Specifically, the extracted multi-dimensional signal features are combined into a state vector s as the input of the reinforcement learning model, and then the model's actions are set. Different signal sources are assigned different quality scores, with values ranging from 0 to 1. Then, a reward function is set for the reinforcement learning model, and the quality score vector of the current time step is , the quality score vector of the previous time step is , then the rate of change of the quality score can be expressed as , extract the preset signal stability threshold in the database ,like When the signal is less than the preset signal stability threshold, additional rewards will be given , on the contrary, there is no reward, and the reward function is expressed as ; The reinforcement learning model calculates each action through forward propagation based on the input state vector s The Q value represents the action taken in state s The model selects the action with the largest Q value as the output, that is, the quality score value corresponding to each signal source is obtained. It should be noted that the number of signal sources is not fixed. The quality score value of the corresponding number is output based on the number of signal sources currently collected by the aircraft in real time.

[0040] Step 3: Dynamic weight adjustment: Consider the weight of each signal as a state variable, construct the observation equation space, and then combine the weight adjustment of the Kalman filter algorithm. Specifically, obtain the number of signal sources currently received and mark it as g, and the weight vector is expressed as Z=[z1,z2,……,z g ], set the state transfer equation to , Expressed as the weight change, the observation equation space is constructed accordingly;

[0041] According to the weight estimate of the previous moment And the state transfer equation calculates the prediction weight, the formula is , and the prediction covariance formula is ,in represents the covariance of the previous moment, Represents the process noise covariance, converts each quality score value into a quality assessment vector and integrates it into the quality assessment result , using the Kalman gain formula Get the Kalman gain value at that moment ,in Represents the observation matrix obtained by mapping the weight state to the observation equation space. Similarly Expressed as the observation noise covariance, the weight estimate is then updated and covariance ,Through continuous iterative prediction and updating steps, the weight of each signal source is adjusted and transmitted to the positioning settlement based on the changes of the signal features collected and extracted in real time.

[0042] Step 4: Positioning and Settlement: Based on the corresponding weights of each signal source, the distance information between multiple signal sources and the aircraft is used to determine the position coordinates of the aircraft through geometric positioning principles, including triangulation and multilateral positioning. Specifically, the free space propagation model is set as the distance measurement formula, and the transmission power, signal wavelength, and receiving power of the broadcast signal receiver, Wi-Fi signal receiver, and Bluetooth signal receiver are obtained, marked as 、 and , substitute the broadcast signal receiver's transmission power, signal wavelength, and receiving power into the formula to calculate Get the broadcast signal distance value d2, and also follow the above steps and formula to get the Wi-Fi signal distance value d3 and the Bluetooth signal distance value d4, that is, , .

[0043] like Figure 4 As shown, with the aircraft as the center point, when three signal sources are obtained, their coordinates are marked as 、 and Calculated based on the triangulation principle Get the two-dimensional position coordinates of the aircraft ; If four signal sources are obtained, their coordinates are marked as 、 、 and , calculated using the principle of multilateration Get the three-dimensional position coordinates of the aircraft .

[0044] For example, taking three-dimensional coordinates as an example, according to the weights Q1, Q2 and Q3 of each signal source obtained by dynamic weight adjustment, the coordinates are 、 and , calculate the three-dimensional coordinates of the aircraft and the corresponding weights Get the final aircraft position .

[0045] like Figure 2 As shown, the second aspect of the present invention provides a high-precision positioning device for an aircraft based on opportunity signal fusion, comprising:

[0046] The opportunity signal acquisition module monitors the signal strength of the aircraft in real time during the aircraft's flight mission. When the signal strength falls below a preset threshold, it generates a signal acquisition instruction and distributes it to multiple signal receiving devices, which then collect the signal.

[0047] The signal feature extraction module extracts features from broadcast signal receivers, Wi-Fi signal receivers, radar signal receivers, and Bluetooth signal receivers, inputs the signal features into a preset reinforcement learning model, and outputs a quality score for each signal source;

[0048] The weight dynamic adjustment module regards the weight of each signal as a state variable, constructs the observation equation space, and then adjusts the weight in real time in combination with the Kalman filter algorithm;

[0049] The positioning settlement module uses the distance information between multiple signal sources and the aircraft based on the corresponding weights of each signal source, and then uses geometric positioning principles, including triangulation and multilateral positioning, to determine the position coordinates of the aircraft, and obtains the final position coordinates based on weighted calculation.

[0050] The above is an illustration of the present invention and should not be considered as limiting thereof. Although several exemplary embodiments of the present invention have been described, it will be readily understood by those skilled in the art that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present invention. Therefore, all such modifications are intended to be included within the scope of the present invention as defined by the claims. It should be understood that the above is an illustration of the present invention and should not be considered as being limited to the specific embodiments disclosed, and modifications to the disclosed embodiments and other embodiments are intended to be included within the scope of the appended claims. The present invention is defined by the claims and their equivalents.

Claims

1. A high-precision aircraft positioning method based on opportunity signal fusion, characterized in that: The steps include: During the flight mission, the monitoring program obtains the signal strength of the current aircraft in real time. When the signal strength is lower than the preset signal strength threshold, it generates a signal collection instruction and distributes it to multiple signal receiving devices, and each signal receiving device collects the signal. According to the extraction of the features of each signal source, the signal features are input into the preset reinforcement learning model, and the quality score of each signal source is output. Specifically, the reinforcement learning model is established using a deep Q network structure, which includes an input layer, multiple hidden layers and an output layer. The input layer receives the extracted multi-dimensional signal features, the hidden layer performs nonlinear transformation and abstraction on the features, and the output layer outputs the quality score of each signal source. Specifically, the extracted multi-dimensional signal features are combined into a state vector s as the input of the reinforcement learning model, and then the model's actions are set. Different signal sources are assigned different quality scores, ranging from 0 to 1. Then, a reward function is set for the reinforcement learning model, and the quality score vector of the current time step is , the quality score vector of the previous time step is , then the rate of change of the quality score is expressed as , extract the preset signal stability threshold in the database ,like When the signal is less than the preset signal stability threshold, additional rewards will be given ; Otherwise, there is no reward, and the reward function is expressed as ; The reinforcement learning model calculates each action through forward propagation based on the input state vector s The Q value represents the action taken in state s The expected cumulative reward, the model selects the action with the largest Q value as the output, that is, the quality score value corresponding to each signal source is obtained; The weight of each signal is regarded as a state variable, the observation equation space is constructed, and the weight is adjusted in real time in combination with the Kalman filter algorithm; According to the corresponding weights of each signal source obtained, the distance information between multiple signal sources and the aircraft is used, and then the position coordinates of the aircraft are determined through geometric positioning principles, including triangulation positioning and multilateral positioning, and the final position coordinates are obtained based on weighted calculation.

2. The method for high-precision aircraft positioning based on opportunity signal fusion according to claim 1, characterized in that: The multi-dimensional signal feature extraction extracts features of a broadcast signal receiver, a Wi-Fi signal receiver, a radar signal receiver, and a Bluetooth signal receiver, specifically: Based on PLL, the carrier phase in the broadcast signal is tracked and the Allan variance formula is used. Get the carrier phase stability value ,in, is the carrier frequency, M is the number of measurements, It is The average frequency within the measurement interval; the logarithmic amplifier converts the RF signal power into a voltage value, which is then sampled by the analog-to-digital converter to calculate the broadcast signal strength sp; the current signal is then demodulated to obtain a real-time modulation signal , extract the standard modulation signal in the database , using the mean square error formula Get the modulation distortion value MSE; Use a spectrum analyzer to detect the interference signal power in the WiFi target channel, and then obtain the target signal power. Divide the interference signal power by the target signal power to obtain the co-channel interference value. By analogy, the radar signal is feature extracted and the time difference between the radar transmitting pulse and the echo pulse is marked as , calculated according to the formula Get the distance d1 between the aircraft and the target radar, where c represents the speed of light; then use the formula to calculate Get the Doppler frequency shift generated by the echo signal , where v represents the relative speed between the aircraft and the radar, λ represents the signal wavelength, and cosθ represents the relative motion angle between the radar and the aircraft; Extract the Bluetooth signal features, mark the start time of the signal acquisition of the first device, and then mark the end time of the acquisition, calculate the time difference between the two marked time points to get the acquisition period, obtain the number of signal acquisitions within the acquisition period and the corresponding signal value marked as ,in , j represents the signal acquisition number, m represents the total number of acquisitions, and the mean value RE of the Bluetooth signal is calculated by the mean value, and then the standard deviation formula is calculated Get the Bluetooth signal standard deviation RQ. A low variance indicates a stable signal.

3. The method for high-precision aircraft positioning based on opportunity signal fusion according to claim 1, characterized in that: The weight of each signal is regarded as a state variable, and the observation equation space is constructed. Then, the weight adjustment of the Kalman filter algorithm is combined, which is as follows: Get the number of signal sources currently received and mark it as g, and the weight vector is represented as Z=[z1,z2,……,z g ], set the state transfer equation to , Expressed as the weight change, the observation equation space is constructed accordingly; According to the weight estimate of the previous moment And the state transfer equation calculates the prediction weight, the formula is , and the prediction covariance formula is ,in represents the covariance of the previous moment, Represents the process noise covariance, converts each quality score value into a quality assessment vector and integrates it into the quality assessment result , using the Kalman gain formula Get the Kalman gain value at that moment ,in Represents the observation matrix obtained by mapping the weight state to the observation equation space. Similarly Expressed as the observation noise covariance, the weight estimate is then updated and covariance ,Through continuous iterative prediction and update steps, according to the changes in the signal ,features collected and extracted in real time, the corresponding weights are ,adjusted according to the number of signal sources and transmitted to the ,positioning settlement.

4. The method for high-precision aircraft positioning based on opportunity signal fusion according to claim 1, characterized in that: The positioning settlement is based on the corresponding weights of each signal source, using the distance information between multiple signal sources and the aircraft, and then using geometric positioning principles, including triangulation and multilateral positioning, to determine the position coordinates of the aircraft. The specific process is as follows: Set the free space propagation model as the distance measurement formula to obtain the transmission power, signal wavelength, and receiving power of the broadcast signal receiver, Wi-Fi signal receiver, and Bluetooth signal receiver, which are marked as 、 and , substitute the broadcast signal receiver's transmission power, signal wavelength, and receiving power into the formula to calculate Obtain the broadcast signal distance value d2, and similarly follow the above steps and formulas to obtain the Wi-Fi signal distance value d3 and the Bluetooth signal distance value d4.

5. The method for high-precision aircraft positioning based on opportunity signal fusion according to claim 4, characterized in that: With the aircraft as the center point, when three signal sources are obtained, their coordinates are marked as 、 and Calculated based on the triangulation principle Get the two-dimensional position coordinates of the aircraft ; If four signal sources are obtained, their coordinates are marked as 、 、 and , calculated using the principle of multilateration Get the three-dimensional position coordinates of the aircraft ; The weights Q1, Q2 and Q3 of each signal source obtained by dynamic weight adjustment are as follows: 、 and , calculate the three-dimensional coordinates of the aircraft and the corresponding weights Get the final aircraft position .

6. The method for high-precision aircraft positioning based on opportunity signal fusion according to claim 1, characterized in that: When an aircraft is performing a flight mission, the monitoring program acquires the signal strength of the aircraft in real time. When the signal strength is lower than a preset signal strength threshold, a signal acquisition instruction is generated and distributed to various signal receiving devices. Specifically, The various signal receiving devices include broadcast signal receivers, Wi-Fi signal receivers, radar signal receivers and Bluetooth signal receivers. The aircraft's various signal receiving devices obtain signal collection instructions and collect various opportunity signals in the surrounding environment in real time from the broadcast signal receivers, Wi-Fi signal receivers, radar signal receivers and Bluetooth signal receivers in turn.

7. The method for high-precision aircraft positioning based on opportunity signal fusion according to claim 6, characterized in that: The opportunity signal acquisition collects data from each signal source, specifically: The medium wave and short wave frequency bands are scanned through the tuning circuit of the broadcast signal receiver. When a broadcast signal is detected, the front-end circuit of the receiver amplifies and filters the broadcast signal to remove noise interference and obtain a processed broadcast signal. The wireless frequency band is then scanned through the Wi-Fi signal receiver. When a hotspot signal is detected, the MAC address of the current hotspot is first identified, the signal strength is detected through the RSSI circuit, and multiple acquisitions are performed on the same hotspot according to the preset interval. The signal strength values are obtained and averaged to obtain the average signal strength. If the average signal strength is greater than the pre-set signal strength threshold, it is determined to be the target signal and sent to the signal feature extraction. Otherwise, the signal source is deleted. By analogy, a radar transmitter is set next to the radar signal receiver, and electromagnetic wave pulses are sent through the radar transmitter. The radar signal receiver receives and amplifies the transmitted echo signal; finally, the Bluetooth signal receiver scans multiple Bluetooth signals according to the preset interval, and obtains the signal strength of each Bluetooth device, arranges the signal strengths, obtains the Bluetooth device with the largest signal strength and marks it as the first device.

8. A high-precision aircraft positioning device based on opportunity signal fusion, characterized in that: The method for high-precision positioning of an aircraft based on opportunity signal fusion is applied to implement any one of claims 1 to 7, comprising: The opportunity signal acquisition module monitors the signal strength of the aircraft in real time during the aircraft's flight mission. When the signal strength falls below a preset threshold, it generates a signal acquisition instruction and distributes it to multiple signal receiving devices, which then collect the signal. The signal feature extraction module extracts features from broadcast signal receivers, Wi-Fi signal receivers, radar signal receivers, and Bluetooth signal receivers, inputs the signal features into a preset reinforcement learning model, and outputs a quality score for each signal source; The weight dynamic adjustment module regards the weight of each signal as a state variable, constructs the observation equation space, and then uses the Kalman filter algorithm to adjust the weight in real time; The positioning settlement module uses the distance information between multiple signal sources and the aircraft based on the corresponding weights of each signal source, and then uses geometric positioning principles, including triangulation and multilateral positioning, to determine the position coordinates of the aircraft, and obtains the final position coordinates based on weighted calculation.

Images