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

Abstract

The invention relates to the field of aircraft high-precision positioning, in particular to an aircraft high-precision positioning method and device based on opportunity signal fusion, and the method comprises the steps: collecting opportunity signals through broadcast, Wi-Fi, radar and a Bluetooth signal receiver when the signal intensity of an aircraft is lower than a preset threshold value; multi-dimensional signal features are extracted, and a deep Q network reinforcement learning model is utilized to output a signal source quality score; a signal weight is regarded as a state variable, and the weight is dynamically adjusted in combination with a Kalman filtering algorithm; and according to the signal source weight and the distance information, determining the position coordinates of the aircraft by adopting triangulation positioning and multilateral positioning, and weighting and calculating the final position. The device comprises an opportunity signal acquisition module, a signal feature extraction module, a weight dynamic adjustment module and a positioning settlement module which respectively execute corresponding functions. According to the method, multiple opportunity signals are fused, and the positioning precision and reliability of the aircraft in a complex environment are improved.

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 the signal, causing multipath effects, seriously interfering with the normal reception of the signal, resulting in a significant decrease in positioning accuracy. In areas with complex electromagnetic environments, such as areas with dense communication base stations or places with strong electromagnetic interference sources, the signal is interfered with and the accuracy of positioning cannot be guaranteed.

[0004] The traditional single positioning method 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 purpose 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, and the specific steps 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 a variety of signal receiving devices. The various signal receiving devices include broadcast signal receivers, Wi-Fi signal receivers, radar signal receivers, and Bluetooth signal receivers. The various signal receiving devices of the aircraft obtain the signal acquisition instruction and collect various opportunity signals in the surrounding environment from the broadcast signal receiver, Wi-Fi signal receiver, radar signal receiver, and Bluetooth signal receiver in real time. The broadcast signal receiver is tuned by the The receiver 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 the same hotspot is collected multiple times according to the preset interval. The signal strength values ​​are obtained and the average is calculated to obtain the signal strength average. If the signal strength average is greater than the pre-set signal strength threshold, it is determined to be a target signal and sent to the signal feature extraction. Otherwise, the signal source is deleted.

[0008] By analogy, a radar transmitter is arranged 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 signals; finally, the Bluetooth signal receiver scans multiple Bluetooth signals according to a preset interval, and obtains the signal strength of each Bluetooth device, arranges the signal strengths, obtains the Bluetooth device with the maximum signal strength and marks it as the first preferred 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 processed broadcast signal based on PLL, and use the Allen variance formula Get the carrier phase stability value β, where f 0 is the carrier frequency, M is the number of measurements, is the average frequency in the i-th measurement interval; the RF signal power is converted into a voltage value through a logarithmic amplifier, and then the broadcast signal strength sp is calculated after sampling by an analog-to-digital converter (ADC); the current signal is demodulated to obtain a real-time modulation signal s demod (t), extract the standard modulation signal s in the database std (t), 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 get the same-frequency 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 t TOF , 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 The Doppler frequency shift Δf generated by the echo signal is obtained, 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;

[0012] Finally, extract the Bluetooth signal features, mark the start time of the signal acquisition of the first preferred device, and then mark the end time of acquisition, and calculate the time difference between the two marked time points to get the acquisition period, obtain the number of signal acquisitions in the acquisition period and the corresponding signal value marked as Rj (j = 1, 2, ..... m), j represents the signal acquisition number, m represents the total number of acquisitions, and perform mean calculation to obtain the Bluetooth signal mean RE, and then perform standard deviation formula calculation 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 action α is set to give different quality scores to different signal sources, and its value is 0-1; then a reward function is set for the reinforcement learning model, and the quality score vector of the current time step is set to γ e , the quality score vector of the previous time step is γ e-1 , then the rate of change of the quality score can be expressed as Extract the preset signal stability threshold Δγ in the database th If Δγ is less than the preset signal stability threshold, an additional reward R is given S , on the contrary, there is no reward, and the reward function is expressed as The reinforcement learning model calculates the Q value of each action α through forward propagation based on the input state vector s. The Q value represents the expected cumulative reward of taking action α under 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, and the corresponding number of quality score values ​​is output based on the number of signal sources currently collected by the aircraft in real time.

[0014] Step 3: Dynamic weight adjustment: 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. Specifically, the number of signal sources currently received is obtained and marked as g. The weight vector is expressed as Z = [z 1 ,z 2 ,……,z g ], set the state transfer equation to Z h =Z h-1 +ΔZ h , ΔZ h Expressed as 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 Where D h-1 represents the covariance of the previous moment, Q h Represents the process noise covariance, converts each quality score value into a quality assessment vector and integrates it into the quality assessment result L p , using the Kalman gain formula Get the Kalman gain value K at this moment h , where H h Represents the observation matrix obtained by mapping the weight state to the observation equation space. Similarly, R h Expressed as the observation noise covariance, the weight estimate is then updated and covariance Through continuous iterative prediction and update steps, the weight of each signal source is adjusted according to the changes in signal characteristics collected and extracted in real time and transmitted to the positioning settlement.

[0016] Step 4: Positioning settlement: According to the corresponding weights of each signal source, the distance information between multiple signal sources and the aircraft is used, and then the position coordinates of the aircraft are determined through the geometric positioning principle, including triangulation positioning and multilateral positioning. Specifically, the free space propagation model is set as the measurement distance 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, which are marked as P respectively. r ,λ,and P t, substitute the transmission power, signal wavelength and receiving power of the broadcast signal receiver into the formula to calculate Get the broadcast signal distance value d2, and also get the Wi-Fi signal distance value d3 and the Bluetooth signal distance value d4 according to the above steps and formulas;

[0017] With the aircraft as the center point, when three signal sources are obtained, their coordinates are marked as (x2, y2), (x3, y3) and (x4, y4) respectively. Get the two-dimensional position coordinates (x, y) of the aircraft; if four signal sources are obtained, mark their coordinates as (x1, y1, z1), (x2, y2, z2), (x3, y3, z3) and (x4, y4, z4) respectively, and calculate using the principle of multilateral positioning Get the three-dimensional position coordinates (x, y, z) of the aircraft;

[0018] According to the weights Q1, Q2 and Q3 of each signal source obtained by dynamic weight adjustment, the three-dimensional coordinates of the aircraft and the corresponding weights are calculated. Get the final aircraft position (x5, y5, z5).

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

[0020] The opportunity signal acquisition module obtains the signal strength of the current aircraft in real time according to the monitoring program during the aircraft's flight mission. When the signal strength is lower than the preset signal strength threshold, it generates a signal acquisition instruction and distributes it to multiple signal receiving devices, and each signal receiving device collects the signal;

[0021] The signal feature extraction module extracts the features of the broadcast signal receiver, Wi-Fi signal receiver, radar signal receiver and Bluetooth signal receiver, inputs the signal features into a preset reinforcement learning model, and outputs the quality score of 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 according to the corresponding weights of each signal source, and then uses the 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 the prior art, the present invention has the following beneficial effects: in the process of signal processing, the present invention extracts multi-dimensional features of the collected signals, and uses the reinforcement learning model to score each signal source, screens out high-quality and stable signals, and in the signal feature extraction stage, comprehensively evaluates the signal quality by calculating the carrier phase stability value and modulation distortion value of the broadcast signal, as well as the co-channel interference value of the Wi-Fi signal. This can reduce the interference of low-quality signals on the positioning results, and further improve the accuracy and stability of positioning;

[0025] Dynamic weight adjustment regards signal weight as a state variable and adjusts the weight in real time in combination with the Kalman filter algorithm; this dynamic adjustment mechanism can automatically optimize the weight distribution of each signal source in the positioning calculation according to the real-time changes in signal characteristics. When the signal quality of a signal source improves, its weight is increased, and vice versa, the weight is reduced, so that the best positioning effect can be achieved under different flight conditions, improving the adaptive ability of the positioning system. 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 The figure is a flowchart of the method steps of the present invention.

[0028] Figure 2 It 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 technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments 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 creative work 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, and the specific steps 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 a variety of signal receiving devices. The various signal receiving devices include broadcast signal receivers, Wi-Fi signal receivers, radar signal receivers, and Bluetooth signal receivers. The various signal receiving devices of the aircraft obtain the signal acquisition instruction and collect various opportunity signals in the surrounding environment from the broadcast signal receiver, Wi-Fi signal receiver, radar signal receiver, and Bluetooth signal receiver in real time. The broadcast signal receiver is tuned by the The receiver 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 the same hotspot is collected multiple times according to the preset interval. The signal strength values ​​are obtained and the average is calculated to obtain the signal strength average. If the signal strength average is greater than the pre-set signal strength threshold, it is determined to be a target signal and sent to the signal feature extraction. Otherwise, the signal source is deleted.

[0034] By analogy, a radar transmitter is arranged 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 signals; finally, the Bluetooth signal receiver scans multiple Bluetooth signals according to a preset interval, and obtains the signal strength of each Bluetooth device, arranges the signal strengths, obtains the Bluetooth device with the maximum signal strength and marks it as the first preferred 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 processed broadcast signal based on PLL, and use the Allen variance formula Get the carrier phase stability value β, where f 0 is the carrier frequency, M is the number of measurements, is the average frequency in the i-th measurement interval; the RF signal power is converted into a voltage value through a logarithmic amplifier, and then the broadcast signal strength sp is calculated after sampling by an analog-to-digital converter (ADC); the current signal is demodulated to obtain a real-time modulation signal s demod (t), extract the standard modulation signal s in the database std (t), 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 get the same-frequency 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 t TOF , 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 The Doppler frequency shift Δf generated by the echo signal is obtained, 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;

[0038] Finally, extract the Bluetooth signal features, mark the start time of the signal acquisition of the first preferred device, and then mark the end time of acquisition, and calculate the time difference between the two marked time points to get the acquisition period, obtain the number of signal acquisitions in the acquisition period and the corresponding signal value marked as Rj (j = 1, 2, ..... m), j represents the signal acquisition number, m represents the total number of acquisitions, and perform mean calculation to obtain the Bluetooth signal mean RE, and then perform standard deviation formula calculation 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 action α is set to give different quality scores to different signal sources, and its value is 0-1; then a reward function is set for the reinforcement learning model, and the quality score vector of the current time step is set to γ e , the quality score vector of the previous time step is γ e-1 , then the rate of change of the quality score can be expressed as Extract the preset signal stability threshold Δγ in the database th If Δγ is less than the preset signal stability threshold, an additional reward R is given S , on the contrary, there is no reward, and the reward function is expressed as The reinforcement learning model calculates the Q value of each action α through forward propagation based on the input state vector s. The Q value represents the expected cumulative reward of taking action α under 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, and the corresponding number of quality score values ​​is output based on the number of signal sources currently collected by the aircraft in real time.

[0040] Step 3: Dynamic weight adjustment: 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. Specifically, the number of signal sources currently received is obtained and marked as g. The weight vector is expressed as Z = [z 1 ,z 2 ,……,z g ], set the state transfer equation to Z h =Z h-1 +ΔZ h , ΔZ h Expressed as 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 Where D h-1 represents the covariance of the previous moment, Q h Represents the process noise covariance, converts each quality score value into a quality assessment vector and integrates it into the quality assessment result L p , using the Kalman gain formula Get the Kalman gain value K at this moment h , where H h Represents the observation matrix obtained by mapping the weight state to the observation equation space. Similarly, R h Expressed as the observation noise covariance, the weight estimate is then updated and covariance Through continuous iterative prediction and update steps, the weight of each signal source is adjusted according to the changes in signal characteristics collected and extracted in real time and transmitted to the positioning settlement.

[0042] Step 4: Positioning settlement: According to the corresponding weights of each signal source, the distance information between multiple signal sources and the aircraft is used, and then the position coordinates of the aircraft are determined through the geometric positioning principle, including triangulation positioning and multilateral positioning. Specifically, the free space propagation model is set as the measurement distance 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, which are marked as P respectively. r ,λ,and P t, substitute the transmission power, signal wavelength and receiving power of the broadcast signal receiver into the formula to calculate Get the broadcast signal distance value d2, and also get the Wi-Fi signal distance value d3 and the Bluetooth signal distance value d4 according to the above steps and formulas, 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 (x2, y2), (x3, y3) and (x4, y4) respectively. Get the two-dimensional position coordinates (x, y) of the aircraft; if four signal sources are obtained, mark their coordinates as (x1, y1, z1), (x2, y2, z2), (x3, y3, z3) and (x4, y4, z4) respectively, and calculate using the principle of multilateral positioning Get the three-dimensional position coordinates (x, y, z) of the aircraft.

[0044] For example, taking the three-dimensional coordinates as an example, according to the weights Q1, Q2 and Q3 of each signal source obtained by dynamic weight adjustment, the three-dimensional coordinates of the aircraft and the corresponding weights are calculated. Get the final aircraft position (x5, y5, z5).

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

[0046] The opportunity signal acquisition module obtains the signal strength of the current aircraft in real time according to the monitoring program during the aircraft's flight mission. When the signal strength is lower than the preset signal strength threshold, it generates a signal acquisition instruction and distributes it to multiple signal receiving devices, and each signal receiving device collects the signal;

[0047] The signal feature extraction module extracts the features of the broadcast signal receiver, Wi-Fi signal receiver, radar signal receiver and Bluetooth signal receiver, inputs the signal features into a preset reinforcement learning model, and outputs the quality score of 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 according to the corresponding weights of each signal source, and then uses the 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 explanation of the present invention and should not be considered as a limitation thereof. Although several exemplary embodiments of the present invention have been described, it will be readily appreciated 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 in the claims. It should be understood that the above is an explanation 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: 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 collection instruction is generated and distributed 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 action α is set to different signal sources, and different quality scores are assigned, whose values ​​are 0-1; then the reward function is set for the reinforcement learning model, and the quality score vector of the current time step is set to γ e , the quality score vector of the previous time step is γ e-1 , then the rate of change of the quality score is expressed as Extract the preset signal stability threshold Δγ in the database th If Δγ is less than the preset signal stability threshold, an additional reward R is given S ; Otherwise, there is no reward, and the reward function is expressed as The reinforcement learning model calculates the Q value of each action α through forward propagation based on the input state vector s. The Q value represents the expected cumulative reward of taking action α under 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; 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, 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 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. The carrier phase stability value β is obtained, where f0 is the carrier frequency and M is the number of measurements. is the average frequency in the i-th measurement interval; the logarithmic amplifier converts the RF signal power into a voltage value, and then samples it through the analog-to-digital converter to calculate the broadcast signal strength sp; then the current signal is demodulated to obtain the real-time modulation signal s demod (t), extract the standard modulation signal s in the database std (t), 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 get the same-frequency 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 t TOF , 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 The Doppler frequency shift Δf generated by the echo signal is obtained, 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 preferred device, and then mark the end time of acquisition, and calculate the time difference between the two marked time points to obtain the acquisition period, obtain the number of signal acquisitions in the acquisition period and the corresponding signal value marked as Rj (j = 1, 2, ..... m), j represents the signal acquisition number, m represents the total number of acquisitions, and perform mean calculation to obtain the Bluetooth signal mean RE, and then perform standard deviation formula calculation The standard deviation RQ of the Bluetooth signal is obtained, and 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, the observation equation space is constructed, and then combined with the weight adjustment of the Kalman filter algorithm, specifically: Get the number of signal sources currently received and mark it as g, and the weight vector is represented by Z = [z1, z2, ..., z g ], set the state transfer equation to Z h =Z h-1 +ΔZ h , ΔZ h Expressed as 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 Where D h-1 represents the covariance of the previous moment, Q h Represents the process noise covariance, converts each quality score value into a quality assessment vector and integrates it into the quality assessment result L p , using the Kalman gain formula Get the Kalman gain value K at this moment h , where H h Represents the observation matrix obtained by mapping the weight state to the observation equation space. Similarly, R h Expressed as the observation noise covariance, the weight estimate is then updated and covariance Through continuous iteration of prediction and update steps, according to the changes in signal characteristics 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 the geometric positioning principle, including triangulation positioning 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 measurement distance formula, and obtain the transmission power, signal wavelength, and receiving power of the broadcast signal receiver, Wi-Fi signal receiver, and Bluetooth signal receiver, marked as P respectively. r , λ and P t , substitute the transmission power, signal wavelength and receiving power of the broadcast signal receiver into the formula to calculate The broadcast signal distance value d2 is obtained, and the Wi-Fi signal distance value d3 and the Bluetooth signal distance value d4 are obtained according to the above steps and formulas.

5. The method for high-precision aircraft positioning based on opportunity signal fusion according to claim 4, characterized in that: The aircraft is taken as the center point. When three signal sources are obtained, their coordinates are marked as (x2, y2), (x3, y3) and (x4, y4) respectively. The triangulation positioning principle is used to calculate Get the two-dimensional position coordinates (x, y) of the aircraft; if four signal sources are obtained, mark their coordinates as (x1, y1, z1), (x2, y2, z2), (x3, y3, z3) and (x4, y4, z4) respectively, and calculate using the principle of multilateral positioning Get the three-dimensional position coordinates (x, y, z) of the aircraft; The weights of each signal source are dynamically adjusted according to the weights and set as Q1, Q2 and Q3, and the three-dimensional coordinates of the aircraft and the corresponding weights are calculated. Get the final aircraft position x5, y5, z5).

6. The method for high-precision aircraft positioning based on opportunity signal fusion according to claim 1, characterized in that: When the aircraft is performing a flight mission, the monitoring program acquires 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 a variety of 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 various signal receiving devices of the aircraft 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; Then the wireless frequency band is scanned by 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 by the RSSI circuit, and the same hotspot is collected multiple times according to the preset interval, and each signal strength value is obtained and averaged to obtain the signal strength average. 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; By analogy, a radar transmitter is arranged 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 signals; finally, the Bluetooth signal receiver scans multiple Bluetooth signals according to a preset interval, and obtains the signal strength of each Bluetooth device, arranges the signal strengths, obtains the Bluetooth device with the maximum signal strength and marks it as the first preferred 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 according to any one of claims 1 to 7 is applied to implement the method, comprising: The opportunity signal acquisition module acquires the signal strength of the current aircraft in real time according to the monitoring program during the aircraft's flight mission. When the signal strength is lower than the preset signal strength threshold, it generates a signal acquisition instruction and distributes it to multiple signal receiving devices, and each signal receiving device collects the signal. The signal feature extraction module extracts the features of the broadcast signal receiver, Wi-Fi signal receiver, radar signal receiver and Bluetooth signal receiver, inputs the signal features into a preset reinforcement learning model, and outputs the quality score of 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 adjusts the weight in real time in combination with the Kalman filter algorithm; The positioning settlement module uses the distance information between multiple signal sources and the aircraft according to the corresponding weights of each signal source, and then uses the geometric positioning principle, including triangulation and multilateral positioning to determine the position coordinates of the aircraft, and obtains the final position coordinates based on weighted calculation.

Images