A method and system for cooperative guidance of unmanned aerial vehicles based on dual-mode communication

Abstract

This application relates to a method and system for collaborative guidance of unmanned aerial vehicles (UAVs) based on dual-mode communication. By introducing dual-mode communication into the guidance and landing device, in environments with good public network coverage, a broadband link is used to ensure high-frequency data interaction with the UAV; in extreme environments without network coverage, it automatically switches to a narrowband satellite communication link to transmit location information to the UAV via short messages. This ensures that even in environments with damaged or dead-zone communication infrastructure, the guidance and landing device can still maintain remote collaborative guidance of the UAV, avoiding the low reliability caused by a single link. This improves the ability of UAVs to receive critical location information in emergency rescue and complex operational scenarios, preventing UAVs from losing contact or failing to land.

Description

Technical Field

[0001] This application relates to the field of drone guidance equipment technology, and in particular to a drone collaborative guidance method and system based on dual-mode communication. Background Technology

[0002] With the widespread application of unmanned aerial vehicles (UAVs) in logistics, emergency rescue, and routine inspection, precise navigation and landing control are crucial for their safe operation. In existing UAV guidance and landing solutions, ground-based guidance and landing equipment typically relies on a single communication network environment. This involves transmitting collected position coordinate data to a backend management platform or the UAV itself via a public cellular network (such as 4G / 5G) or a specific ground-based line-of-sight microwave data link to guide and control the UAV's flight.

[0003] However, in practical applications, ground-based guidance and landing equipment that relies on a single communication environment will lose its data transmission link and become offline when faced with blind spots in public network base station coverage or damaged communication infrastructure in complex geographical environments or specific emergency rescue scenarios. Due to the lack of emergency cross-network communication means, the guidance and landing equipment in the offline state cannot send trigger signals and positioning coordinate data to the back-end management platform or the drone, causing the collaborative guidance and precise landing functions of the drone system to completely fail, which greatly limits the reliability and application scope of unmanned aerial vehicles in network-free or extreme environments. Summary of the Invention

[0004] This application aims to address the technical problem that existing drone guidance devices rely on a single communication link, which leads to guidance failure and drones being unable to land accurately due to link disconnection in complex geographical environments or scenarios where public network base stations are damaged. It proposes a drone collaborative guidance method and system based on dual-mode communication.

[0005] This application provides a UAV cooperative guidance method based on dual-mode communication, applied to a guidance and stopping device, including: Real-time detection of the connection status of broadband communication links, and determination of whether the connection status meets preset connectivity conditions; If the connectivity conditions are met, it is determined to be in network mode, the first coordinate data is obtained, and the first coordinate data is sent to the back-end management platform through the broadband communication link so that the back-end management platform can guide the drone to fly based on the first coordinate data; If the connectivity conditions are not met, the system is determined to be in a no-network mode. The second coordinate data is then acquired and sent to the backend management platform via a narrowband satellite communication link in the form of a short message, so that the backend management platform can guide the drone's flight based on the second coordinate data.

[0006] In one embodiment, the navigation and stopping device includes an RTK positioning module. The step of acquiring second coordinate data and sending the second coordinate data to the backend management platform in the form of short messages via a narrowband satellite communication link includes: Within a preset fitting time window, the target's three-dimensional coordinate time series in the current area is acquired at high frequency based on the RTK positioning module. The three-dimensional coordinate time series of the target is fitted using Chebyshev polynomials. The coefficients of the zero-degree terms in the fitting results are extracted as the reference coordinates representing the absolute position, and the coefficients of the non-zero-degree terms in the fitting results are extracted as the trajectory coefficients representing the relative fluctuations. The reference coordinates are encoded using unsigned integers and a dynamic bit-width allocation algorithm is performed on the trajectory coefficients. The encoded reference coordinates and the trajectory coefficients quantized by the dynamic bit-width allocation algorithm are used together as the second coordinate data. The second coordinate data is combined with the reference timestamp and sampling period features corresponding to the fitting time window to encapsulate and generate a custom short message that meets the single transmission payload limit; Determine whether the current conditions for satellite launch are met based on the locally configured communication frequency smoother; If the conditions are met, the custom short message will be sent to the backend management platform via the narrowband satellite communication link.

[0007] In one embodiment, the step of performing polynomial fitting on the target three-dimensional coordinate time series using Chebyshev polynomials, extracting the zero-degree coefficients from the fitting result as reference coordinates representing the absolute position, and extracting the non-zero-degree coefficients from the fitting result as trajectory coefficients representing relative fluctuations includes: Obtain each physical sampling time corresponding to the fitting time window, and map each physical sampling time to an orthogonal domain normalized variable; Based on the normalized variables, calculate the Chebyshev fitting coefficients of each order; The coefficients corresponding to the zeroth order in the obtained fitting coefficients are defined as the reference coordinates, and the coefficients corresponding to the first to m orders are defined as the trajectory coefficients, where m is the fitting order.

[0008] In one embodiment, after the step of defining the coefficients corresponding to the first to m terms as the trajectory coefficients, the method further includes: Based on the reference coordinates and the trajectory coefficients, the fitted coordinates corresponding to each of the physical sampling times are reconstructed, and the fitting residual between the fitted coordinates and the target three-dimensional coordinate time series is calculated; Determine whether the fitting residuals meet the preset accuracy conditions. If not, dynamically adjust the fitting order m or the length of the fitting time window based on the fitting residuals.

[0009] In one embodiment, the step of encoding the reference coordinates using unsigned integers and performing a dynamic bit-width allocation algorithm on the trajectory coefficients, and then using the encoded reference coordinates and the trajectory coefficients quantized by the dynamic bit-width allocation algorithm together as the second coordinate data, includes: Based on a preset scaling factor, the floating-point reference coordinates are converted into unsigned integers of a specified bit width to obtain the absolute reference code value; Obtain the fitting order, and assign corresponding quantization bit widths to each order coefficient in the trajectory coefficients based on the fitting order; Based on the allocated quantization bit width, discretization quantization is performed on the trajectory coefficients of each order to obtain the relative coefficient encoding matrix; The absolute reference code value and the relative coefficient code matrix are concatenated into a bitstream to generate the second coordinate data.

[0010] In one embodiment, the step of converting the floating-point reference coordinates into an unsigned integer of a specified bit width based on a preset scaling factor to obtain an absolute reference code value includes: The values ​​of the reference coordinates on the three spatial axes are extracted respectively, and then multiplied by the scaling factor to map them to the integer domain; The mapped axis values ​​are stored in fixed-point format using 24-bit unsigned integer pairs to obtain the absolute reference code value.

[0011] In one embodiment, the step of assigning corresponding quantization bit widths to each order coefficient in the trajectory coefficients based on the fitting order includes: Based on the fitting order, obtain the order number k corresponding to each order coefficient in the trajectory coefficients, where 1≤k≤m, and m is the fitting order; According to the order number k in ascending order, the corresponding trajectory coefficients are assigned quantization bit widths in a decreasing trend; Set the upper limit of the total bit width of the relative coefficient encoding matrix corresponding to the trajectory coefficients of each order; If the fitting order reaches a preset order threshold, the load constraint truncation logic is triggered: according to the priority order k from large to small, the bit width of the corresponding trajectory coefficients is compressed in turn, or the items whose order number k exceeds the preset truncation number are directly discarded, until the total bit width of the trajectory coefficients of all orders is less than or equal to the upper limit of the total bit width.

[0012] In one embodiment, after the step of sending the second coordinate data to the background management platform in the form of a short message via a narrowband satellite communication link, the background management platform receives the second coordinate data sent in the form of the short message, performs inverse quantization analysis on it, and extracts the reference coordinates and trajectory coefficients. The back-end management platform uses Chebyshev polynomials to perform continuous interpolation reconstruction based on the reference coordinates and the trajectory coefficients to obtain a continuous three-dimensional trajectory curve. The back-end management platform predicts the target heading point based on the continuous three-dimensional trajectory curve and encapsulates it into UAV control commands, which are then sent to the UAV's flight control system.

[0013] This application also provides a drone cooperative guidance system based on dual-mode communication, including: a guidance and stopping device, a back-end management platform, and a drone; The control device is used to detect the connection status of its own broadband communication link in real time and determine whether the connection status meets the preset connectivity conditions. It is also used to determine that the network is available if the connectivity conditions are met, obtain the first coordinate data, and send the first coordinate data to the background management platform through the broadband communication link; The back-end management platform is used to guide the drone's flight based on the first coordinate data; The navigation device is also used to determine the network-free mode if the connectivity conditions are not met, obtain the second coordinate data, and send the second coordinate data to the background management platform in the form of a short message through a narrowband satellite communication link; The back-end management platform is used to guide the drone's flight based on the second coordinate data.

[0014] In one embodiment, the stop device includes an RTK positioning module, a control module, and a communication frequency smoother; Within a preset fitting time window, the RTK positioning module is used to collect the target three-dimensional coordinate time series of the current area at high frequency, and send the target three-dimensional coordinate time series to the control module; The control module is used to perform polynomial fitting on the target three-dimensional coordinate time series using Chebyshev polynomials, and extract the zero-degree term coefficients in the fitting results as the reference coordinates representing the absolute position, and extract the non-zero-degree term coefficients in the fitting results as the trajectory coefficients representing the relative fluctuations. The control module is also used to encode the reference coordinates using unsigned integers and to perform a dynamic bit width allocation algorithm on the trajectory coefficients, and to use the encoded reference coordinates and the trajectory coefficients quantized by the dynamic bit width allocation algorithm together as the second coordinate data. The control module is also used to combine the second coordinate data with the reference timestamp and sampling period features corresponding to the fitting time window, and encapsulate and generate a custom short message that meets the single transmission payload limit. The communication frequency smoother is used to determine whether the current conditions for satellite launch are met. If the conditions are met, the control module is used to send the custom short message to the backend management platform through the narrowband satellite communication link.

[0015] Beneficial effects: This application introduces dual-mode communication into the guidance and landing device. In environments with good public network coverage, it utilizes a broadband link to ensure high-frequency data interaction with the UAV; in extreme environments without network coverage, it automatically switches to a narrowband satellite communication link to transmit location information to the UAV via short messages. This avoids the low reliability caused by a single link and ensures that even in environments with damaged or dead-zone communication infrastructure, the guidance and landing device can still maintain remote collaborative guidance for the UAV. This improves the ability of UAVs to receive critical location information in emergency rescue and complex operational scenarios, preventing UAVs from losing contact or failing to land. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 This is a flowchart illustrating an embodiment of the UAV cooperative guidance method based on dual-mode communication according to this application.

[0017] Figure 2 This is a flowchart illustrating one step of a UAV cooperative guidance method based on dual-mode communication according to an embodiment of this application.

[0018] Figure 3 This is a schematic diagram of the structure of a drone cooperative guidance system based on dual-mode communication according to another embodiment of this application. Detailed Implementation

[0019] The terms "first," "second," and "third," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the term "comprising," and any variations thereof, is intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or modules is not limited to the listed steps or modules, but may optionally include steps or modules not listed, or may optionally include other steps or modules inherent to these processes, methods, products, or apparatuses.

[0020] Please see Figure 1-3 One embodiment of this application provides a UAV cooperative guidance method based on dual-mode communication, applied to a navigation and stopping device. The method includes: S1. Real-time detection of the connection status of the broadband communication link, and determination of whether the connection status meets the preset connectivity conditions; S2. If the connectivity condition is met, it is determined to be in network mode. First coordinate data is obtained and sent to the back-end management platform through the broadband communication link so that the back-end management platform can guide the drone to fly based on the first coordinate data. S3. If the connectivity conditions are not met, the system is determined to be in a no-network mode. The second coordinate data is obtained and sent to the back-end management platform in the form of a short message via a narrowband satellite communication link, so that the back-end management platform can guide the UAV to fly based on the second coordinate data.

[0021] As described above, this application can be applied to remote areas, mountainous regions, or scenarios with intermittent network signals or no network coverage. This embodiment is applied to a ground-based navigation and landing device, which in some optional scenarios can also be called a ground-based positioning and guidance device or a ground positioning and guidance terminal. Its function is to collect high-precision location information and communicate collaboratively with external systems. This embodiment sets up a dual-mode communication method. First, it detects the connection status of the broadband access link, such as detecting the 4G network signal strength in the current area. When the 4G network signal strength is greater than a preset threshold, the connectivity condition is met. At this time, the RTK module built into the navigation and landing device obtains accurate first coordinate data. The first coordinate data is usually high-frequency RTK raw coordinates. The navigation and landing device uses transmission protocols such as TCP / UDP to send the first coordinate data to the backend management platform through the broadband communication link, so that the backend management platform can guide the drone's flight based on the first coordinate data. The drone can perform high-precision flight and automatic landing according to the first coordinate data.

[0022] When the 4G network signal strength is not greater than a preset threshold, the connectivity condition is not met, and the current mode is determined to be without network. At this time, the guidance and landing device automatically cuts off the broadband uplink logic and calls the narrowband satellite communication link, such as the Beidou short message system. Due to the extremely small single-pass payload of the narrowband link and the physical limitation of communication frequency, the RTK module obtains accurate second coordinate data and encapsulates it into a short message of a specific format. This short message is transmitted via the satellite link and forwarded by the satellite ground station to the back-end management platform in the network environment. The back-end management platform parses and reconstructs the second coordinate data received by the narrowband to restore the second coordinate data of the guidance and landing device, and then continues to remotely control the drone through the broadband network. This realizes continuous collaborative guidance even when the guidance and landing device is in a network-free state, effectively improving the reliability of the drone guidance equipment. It ensures that the guidance and landing device can still maintain remote collaborative guidance of the drone in environments where communication infrastructure is damaged or in blind spots, improving the accurate positioning in emergency rescue and complex operation scenarios, ensuring that the drone can receive key location information and trigger signals, and preventing the drone from losing contact or failing to land.

[0023] It should be noted that in the offline mode, the navigation device of this embodiment can send the second coordinate data to the background management platform at preset intervals. In addition, the navigation device is equipped with a button and a short message communication module. Users can also activate the RTK module by pressing the button to send the second coordinate data to the background management platform. In addition to the 4G network signal strength mentioned above, the connectivity condition determination criteria of this embodiment can also be the heartbeat packet loss rate or network latency.

[0024] More preferably, the navigation and landing device in this embodiment is also equipped with an "SOS" emergency button. When the "SOS" emergency button is pressed, the navigation and landing device uploads the current location information via a short message. After receiving the SOS command, the background management platform sends the current location information to the drone, activates the preset flight program and emergency landing program, or adjusts the flight attitude.

[0025] In one embodiment, the navigation and stopping device includes an RTK positioning module. Step S3, which involves acquiring second coordinate data and sending the second coordinate data to the backend management platform via a narrowband satellite communication link in the form of short messages, includes: S31. Within a preset fitting time window, the target three-dimensional coordinate time series of the current area is collected at high frequency based on the RTK positioning module. As described above, within a preset fitting time window, the guided stop device uses its built-in RTK positioning module to collect the longitude, latitude, and altitude information of the current area in real time at a preset frequency, such as 1Hz or 5Hz, and converts it into X-axis, Y-axis, and Z-axis coordinate components corresponding to the three spatial axes, constructing a target three-dimensional coordinate time series. The target three-dimensional coordinate time series contains N discrete sampling points, each corresponding to a physical sampling time.

[0026] S32. Use Chebyshev polynomials to perform polynomial fitting on the target three-dimensional coordinate time series, extract the zero-degree term coefficients in the fitting results as the reference coordinates representing the absolute position, and extract the non-zero-degree term coefficients in the fitting results as the trajectory coefficients representing the relative fluctuations. As described above, the guiding stop instrument extracts the start and end times of the fitting time window, converts each physical sampling time into orthogonal domain normalized variables using a mapping formula, and then uses an m-order Chebyshev polynomial to perform least-squares fitting on the target three-dimensional coordinate time series. In the solution, the coefficients corresponding to the zero-degree term of the polynomial are extracted as the reference coordinates representing the physical position of the spatial centroid within the current fitting period, i.e., the reference coordinates of the absolute position (denoted as C0), and the coefficients corresponding to the first to m-degree terms are extracted as trajectory coefficients representing the relative motion attitude of the trajectory (denoted as C1...C1). m This enables dimensionality reduction and compression of the second coordinate data, ensuring that the high-precision continuous trajectory features at the RTK level can be completely stored in a short message payload with narrow bandwidth, without data overflow or out-of-bounds occurrence.

[0027] S33. The reference coordinates are encoded using unsigned integers and a dynamic bit width allocation algorithm is performed on the trajectory coefficients. The encoded reference coordinates and the trajectory coefficients quantized by the dynamic bit width allocation algorithm are used together as the second coordinate data. As described above, the reference coordinates are encoded using 24-bit unsigned integers combined with a preset scaling factor to generate absolute reference code values. For the trajectory coefficients from first order to m order, the processor executes a dynamic bit-width allocation algorithm: based on the physical characteristic that the higher the order of the coefficients, the smaller their numerical envelope, a high bit width (e.g., 16 bits) is allocated to lower order terms, and a low bit width (e.g., 6 to 10 bits) is allocated to higher order terms. Furthermore, by truncating the mantissa of higher order minima, the trajectory coefficients of all orders are compressed into a relative coefficient encoding matrix. The absolute reference code value and the relative coefficient encoding matrix together constitute the second coordinate data.

[0028] S34. Combine the second coordinate data with the reference timestamp and sampling period features corresponding to the fitting time window, and encapsulate them to generate a custom short message that meets the single transmission payload limit. As described above, the navigation and stopping device combines the second coordinate data with the reference timestamp corresponding to the fitted time window, the sampling frequency feature bits, and the device identifier to perform bitstream splicing. During the encapsulation process, the message type is defined by the preset frame header identifier bits, and the cyclic redundancy check code is calculated to finally generate a custom short message that meets the single transmission payload limit of the narrowband satellite communication link (e.g., no more than 78 bytes).

[0029] It should be noted that the reference timestamp can be a time tag used as a reference, such as the start time in the physical sampling time. The sampling frequency feature bit uses one byte to represent the above-mentioned preset frequency, for example, binary 00 represents 1Hz, and 01 represents 5Hz. The device identifier is the device ID of the guide and stop device corresponding to the second coordinate data. In actual field rescue or port operations, there may be drone formations in the sky and multiple guide and stop devices on the ground. Therefore, when encapsulating short messages, the device identifier needs to be added to the header so that the backend management platform can accurately match the guide and stop device with the corresponding drone.

[0030] S35. Determine whether the current conditions for satellite launch are met based on the locally configured communication frequency smoother; if so, send the custom short message to the background management platform through the narrowband satellite communication link.

[0031] As described above, the stop device invokes a locally configured communication frequency smoother to monitor the hardware transmission status of the narrowband satellite communication link. The communication frequency smoother maintains a transmission counting cycle based on the minimum inbound time interval (e.g., 60 seconds) of the narrowband satellite communication link. If the current time falls within a prohibited transmission window, the encapsulated custom short message is stored in the transmission queue for peak smoothing; if the satellite launch conditions are met, a hardware transmission command is triggered.

[0032] In one embodiment, step S32, which involves using Chebyshev polynomials to perform polynomial fitting on the target's three-dimensional coordinate time series, extracting the zero-degree coefficients from the fitting result as reference coordinates representing the absolute position, and extracting the non-zero-degree coefficients from the fitting result as trajectory coefficients representing relative fluctuations, includes: S321. Obtain each physical sampling time corresponding to the fitting time window, and map each physical sampling time to an orthogonal domain normalized variable. As described above, the target's three-dimensional coordinate time series is extracted, and then orthogonally decomposed in the spatial dimension to obtain independent X-axis, Y-axis, and Z-axis coordinate component sequences. For any coordinate component sequence (X-axis, Y-axis, Z-axis), the start time T0 and end time T of the fitting time window are extracted. N And the physical sampling times T within the fitting time window. i(Where i = 1, 2, ..., N, and N is the total number of sampling points), since the basis functions of the Chebyshev polynomial are orthogonal on the interval [-1, 1], in order to eliminate the influence of different dimensions on the fitting accuracy and ensure the optimality of the polynomial approximation, the guide stop instrument uses the following linear mapping formula to normalize each physical sampling time: ; Wherein, the t i For normalized variables, and t i ∈[-1, 1].

[0033] S322. Calculate the Chebyshev fitting coefficients of each order based on the normalized variables; As described above, the decomposed coordinate sequence of each axis is fitted and solved using a Chebyshev polynomial of the first kind as a basis function. The Chebyshev polynomial T... k (t) follows the following recurrence relation: T0(t)=1 (t=1; T1(t) = t (t) = t; (k≥2).

[0034] Where k is the order.

[0035] The stop instrument targets the X-axis coordinate component sequence, Y-axis coordinate component sequence, and Z-axis coordinate component sequence respectively, combined with the normalized variable t. i The discrete orthogonal approximation algorithm is used to calculate the fitting coefficients for each axis at each order. Taking the X-axis as an example, its k-th order fitting coefficient... The calculation formula is as follows: ; Where X(t) i ) represents the X-axis coordinate component sequence at t i The coordinate value corresponding to time T k (t i ) represents the basis function of the k-th order Chebyshev polynomial in the normalized variable t. i The scalar function values ​​at each point. Through the above calculations, the coefficient vectors [C0, C1,..., C] corresponding to the three spatial axes are obtained. m ], where m is the preset or adaptively adjusted fitting order.

[0036] S323. Define the coefficients of the zeroth-order terms in the obtained fitting coefficients as the reference coordinates, and define the coefficients of the first to m-order terms as the trajectory coefficients, where m is the fitting order.

[0037] After obtaining the coefficient vectors for each axis, the guide stop instrument classifies the fitted coefficients according to their physical function based on the order number k: First, extract the coefficients of the zeroth-order terms corresponding to k=0 from the coefficient vectors of each axis, and combine them into a vector. The reference coordinates are defined as follows. Then, the remaining coefficients with indices k=1 to m are extracted from the coefficient vectors corresponding to the X, Y, and Z axes respectively to construct an m-row, 3-column trajectory coefficient matrix. Each element in this matrix represents the dynamic deviation of the trajectory relative to the reference coordinates at various orders. The trajectory coefficient matrix is ​​used as the trajectory coefficients, thus achieving dimensionality reduction of the three-dimensional coordinate data.

[0038] In one embodiment, after step S323, which defines the coefficients corresponding to terms of order one to m as the trajectory coefficients, the method further includes: S324. Based on the reference coordinates and the trajectory coefficients, reconstruct the fitting coordinates corresponding to each of the physical sampling times, and calculate the fitting residual between the fitting coordinates and the target three-dimensional coordinate time series; S325. Determine whether the fitting residual meets the preset accuracy condition. If not, dynamically adjust the fitting order m or the length of the fitting time window based on the fitting residual.

[0039] As mentioned above, the fitted coordinates along the X-axis Taking reconstruction as an example, the calculation formula is: ; in, This represents the reference coordinates of the X-axis within the fitting time window under the zero-order term coefficient. This represents the trajectory coefficient of the kth order corresponding to the X-axis.

[0040] Similarly, the fitted Y-axis coordinates are obtained. Fitted coordinates along the Z-axis .

[0041] Extract the target's three-dimensional coordinate time series (X(T) from the original buffer) i ), Y(T i ), Z(T i The process involves calculating the spatial three-dimensional Euclidean distance between each point and the fitted coordinates, using the spatial three-dimensional Euclidean distance as the single-point position deviation, traversing all N sampling points within the fitting time window, extracting the maximum value of the single-point position deviation (i.e., the maximum envelope error) or calculating the root mean square error, and defining it as the overall fitting residual for the current fitting period.

[0042] Determine whether the fitting residual meets the preset accuracy condition. For example, the preset accuracy threshold is 0.05m. Compare the fitting residual with 0.05m. If the fitting residual is ≤0.05m, it is determined that the current fitting residual meets the preset accuracy condition, indicating that the current fitting order m can truly restore the spatial trajectory under a very high compression ratio. Output the currently calculated reference coordinates and trajectory coefficients according to the set process.

[0043] If the error is greater than 0.05m, the current fitting residual is determined to be insufficient to meet the accuracy requirements. This indicates that the current guiding and stopping instrument is in a high-frequency relative motion state (e.g., affected by wind loads or severe turbulence of its mounting platform, resulting in significant changes in trajectory curvature). The current fitting order cannot accurately approximate the complex real trajectory, thus resulting in underfitting. In this case, the adaptive correction logic is triggered, and the following progressive adjustment is performed: ① Determine whether the current fitting order m is less than the maximum order threshold allowed by the system hardware. max If m < m max Then let the fitting order m = m + 1, and re-execute the system decomposition (S322) and residual verification (S324-S325) using the updated order.

[0044] ②If the current fitting order m has reached the highest order threshold m max Furthermore, the fitting residuals still exceed the standard, indicating that the trajectory curvature within the current time window is too large. Therefore, the length of the fitting time window is shortened according to the preset reduction step size (such as reducing the original window length by 30%), thereby reducing the data time span and geometric complexity of a single fitting, and the sampling and fitting process is retried based on the shortened time window (S31 and S32).

[0045] This enables adaptive adjustment of parameters, ensuring that even in complex and ever-changing maneuvering environments, the issued trajectory coefficients can still faithfully reproduce the dynamic position deviation set of the navigation instrument relative to its spatial centroid reference coordinates within the fitting time window. This allows for the subsequent reconstruction of the target's continuous three-dimensional trajectory curve at each physical sampling moment, and then the calculation of the predicted heading for the next second.

[0046] In one embodiment, step S33, which involves encoding the reference coordinates using unsigned integers and performing a dynamic bit-width allocation algorithm on the trajectory coefficients, and then using the encoded reference coordinates and the trajectory coefficients quantized by the dynamic bit-width allocation algorithm together as the second coordinate data, includes: S331. Based on a preset scaling factor, the floating-point reference coordinates are converted into unsigned integers of a specified bit width to obtain an absolute reference code value. Specifically, the values ​​of the reference coordinates on the three spatial axes are extracted respectively, and multiplied by the scaling factor to map to the integer field; the mapped axis values ​​are stored in fixed-point format using 24-bit unsigned integers to obtain absolute reference code values, so that the reference coordinates achieve centimeter-level absolute positioning accuracy.

[0047] For example, reference coordinates The original storage format is double-precision or single-precision floating-point. Since it represents absolute geographical location, in order to eliminate redundant sign and exponent bits in the floating-point structure, the guide station extracts the values ​​of the reference coordinates on the three spatial axes respectively, and multiplies them uniformly by a preset scaling factor (e.g., set to 10). 2 Or 10 3 This is done by mapping the values ​​to a centimeter or millimeter coordinate system. The non-integer portion is truncated and mapped to the positive integer domain. Subsequently, the mapped axis values ​​are stored in fixed-point format using 24-bit unsigned integers. The fixed-point formatted bit sequences of the three axis components are concatenated to generate an absolute reference code value with a total width of 72 bits. Since the maximum digital capacity of a 24-bit system is 2^32 bits... 24 -1, in matching the centimeter-level scaling factor (10) 2 When used, its maximum representable physical range is approximately 167 kilometers, which can cover the operating radius of a typical UAV and avoids the unnecessary bandwidth redundancy of 8 bits / axis caused by using standard 32-bit integer shaping. The three spatial axes occupy a total of 72 bits, or 9 bytes, which can compress the load overhead of the absolute coordinates of the navigation system to an extremely low level.

[0048] S332. Obtain the fitting order, and assign corresponding quantization bit widths to each order coefficient in the trajectory coefficients based on the fitting order; Specifically, based on the fitting order, the order index k corresponding to each order coefficient in the trajectory coefficients is obtained, where 1≤k≤m, and m is the fitting order; according to the order index k in ascending order, a decreasing quantization bit width is assigned to the corresponding trajectory coefficients; the upper limit of the total bit width of the relative coefficient encoding matrix corresponding to each order trajectory coefficient is set; if the fitting order reaches a preset order threshold, the load constraint truncation logic is triggered: according to the priority order of the order index k in descending order, the bit width of the corresponding trajectory coefficients is compressed sequentially, or items with the order index k exceeding the preset truncation index are directly discarded, until the total bit width of the trajectory coefficients of all orders is less than or equal to the upper limit of the total bit width.

[0049] For example, the current fitting order m is extracted, and the order index k corresponding to each order coefficient in the trajectory coefficient matrix is ​​obtained (where 1≤k≤m). Since low-order coefficients dominate the general direction of the trajectory (large numerical envelope), and high-order coefficients dominate the micro-high frequency fluctuations (small numerical envelope), the corresponding trajectory coefficients can be assigned an initial quantization bit width in a non-increasing trend according to the order index k in ascending order (for example, 16 bits for the first term, 12 bits for the second term, and 8 bits for the third term).

[0050] Further, the system reads the preset total bit width upper limit (this upper limit is calculated based on the single payload limit and frame header overhead of the narrowband satellite communication link). If the fitting order m reaches the preset order threshold, causing the total bit width calculated according to the initial quantization bit width to exceed the upper limit, the payload constraint truncation logic is triggered: compression actions are performed sequentially according to the priority order (i.e., reverse order) of the order index k from largest to smallest. The compression actions include: compressing the mantissa bit width of the corresponding higher-order coefficients through shift operations, or directly discarding higher-order frequency terms with order index k greater than the preset truncation index, until the sum of the quantization bit widths of all retained order trajectory coefficients is strictly less than or equal to the total bit width upper limit.

[0051] S333. Based on the allocated quantization bit width, perform discretization quantization on the trajectory coefficients of each order to obtain the relative coefficient encoding matrix; As described above, after completing the bit width allocation and constraint truncation, corresponding quantization intervals and quantization step sizes are constructed based on the final quantization bit widths determined for each order. Subsequently, discretization quantization is performed on the trajectory coefficients of each order of floating-point type, mapping their values ​​to discrete integers within the corresponding bit width interval. All quantized order coefficients are then reorganized according to the original array arrangement logic to obtain a relative coefficient encoding matrix composed of discrete integers.

[0052] S334. The absolute reference code value and the relative coefficient code matrix are concatenated into a bit stream to generate the second coordinate data.

[0053] As described above, the stop device constructs a contiguous buffer address in local memory, uses the 72-bit absolute reference code value as the payload header, and sequentially fills in the relative coefficient code matrix in ascending order of order k using a compact bit-level arrangement, performing seamless bitstream splicing. During this process, the operating system's default byte alignment mechanism is forcibly disabled to eliminate data gaps. The resulting continuous binary sequence after splicing is the second coordinate data, directly used as the net payload data (second coordinate data) for subsequent custom short message encapsulation.

[0054] In one embodiment, after the step of sending the second coordinate data to the backend management platform via a narrowband satellite communication link in the form of short messages, the method further includes: S301. The background management platform receives the second coordinate data sent in the form of the short message, performs inverse quantization analysis on it, and extracts the reference coordinates and trajectory coefficients. As described above, the backend management platform receives custom short messages from the navigation and stopping device via the satellite ground gateway station and public network routing. The backend management platform unpacks the message according to a preset frame structure protocol and extracts the payload data, including: absolute reference code value, relative coefficient code matrix, and corresponding reference timestamp and sampling frequency feature bits.

[0055] Subsequently, the back-end management platform executes inverse quantization mapping logic that is completely symmetrical with that of the stop instrument: ① For the absolute reference code value, truncate it bit by bit according to the 24-bit / axis width of the three spatial axes, and divide it by the same preset scaling factor to restore it from the integer field to the floating-point reference coordinate. .

[0056] ② For the relative coefficient encoding matrix, read the system's pre-configured dynamic bit width allocation dictionary, and according to the allocated bit width and quantization step size corresponding to each order number k, perform reverse segmentation and dediscretization calculation on the continuously arranged bits in the matrix to restore the trajectory coefficients of each order composed of floating-point values.

[0057] S302. Based on the reference coordinates and the trajectory coefficients, continuous interpolation reconstruction is performed using Chebyshev polynomials to obtain a continuous three-dimensional trajectory curve. As mentioned above, the backend management platform parses the baseline timestamp T0 and the sampling frequency feature bits to calculate the physical time window [T0, T] during the fitting process on the guide stop instrument side. N To achieve high-smoothness spatial interpolation, for any continuous physical time T within the fitting time window (i.e., not limited to the original discrete sampling points, T∈[T0,T...]),... N The back-end management platform uses the following formula: ; Where C0 represents the reference coordinates obtained after inverse quantization analysis. Its corresponding zero-order coefficient represents the spatial centroid (absolute geographical location) of the trajectory within the fitted time window; k is the order number, C k The k-th order trajectory coefficients obtained after dequantization analysis T k (t) represents the scalar function value of the k-th order Chebyshev polynomial basis function at the normalized variable t, which serves as a weighting factor to control the trajectory coefficients C of each order.k The weight of the influence of T on the three-dimensional coordinate vector P(T) at the current moment.

[0058] Since the input variable T is continuously differentiable within its domain, the above reconstruction process mathematically generates a set of smooth analytical functions, thereby expanding and reconstructing the discrete coefficients with finite capacity into a continuous three-dimensional trajectory curve, fundamentally eliminating the step and jump in location data caused by the low-frequency characteristics of narrowband communication (such as once per minute).

[0059] S303. Based on the continuous three-dimensional trajectory curve, predict the target heading point and encapsulate it as a UAV control command to send to the UAV's flight control system.

[0060] As described above, the background management platform extracts the end boundary state vector of the reconstructed continuous three-dimensional trajectory curve (covering end position, instantaneous linear velocity and instantaneous acceleration information, wherein the end position is determined by the zeroth order value of the three-dimensional coordinate vector P(T) formula, the instantaneous linear velocity is determined by the first time derivative of the three-dimensional coordinate vector P(T) formula, and the instantaneous acceleration is determined by the second time derivative of the three-dimensional coordinate vector P(T) formula), and calculates the communication link delay generated from the transmission of the short message to the platform's reception and parsing.

[0061] Combining the communication link latency with the system's preset control time, the backend management platform uses a Kalman filter algorithm or a kinematic smoothing extrapolation model to calculate the predicted target heading coordinates for the next control cycle. Subsequently, based on the UAV flight control underlying communication protocol (such as the MAVLink protocol), the backend management platform encapsulates the calculated target heading coordinates into a UAV control command containing the target's latitude, longitude, altitude, and desired yaw angle fields. Finally, the backend management platform sends this command to the UAV flight control system in the air via a normal broadband downlink communication link (such as a 5G / 4G public network). This allows for accurate prediction of the future real spatial position of the navigation system even under severe satellite communication latency, thereby guiding the UAV and reducing its flight lag.

[0062] It is understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0063] Based on the same inventive concept, this application also provides a system for implementing the above-described dual-mode communication-based UAV cooperative guidance method. The solution provided by this system is similar to the implementation described in the above method; therefore, the specific limitations in one or more system embodiments provided below can be found in the limitations of the method described above, and will not be repeated here.

[0064] In one exemplary embodiment, such as Figure 3 As shown, a drone collaborative guidance system based on dual-mode communication is provided, including: a guidance and stopping device, a background management platform, and a drone.

[0065] The above system corresponds to the aforementioned method embodiments and can implement the corresponding method steps. Its implementation principle and technical effect are similar, and will not be repeated here.

[0066] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein, and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A UAV cooperative guidance method based on dual-mode communication, applied to a navigation and stopping device, characterized in that, include: Real-time detection of the connection status of broadband communication links, and determination of whether the connection status meets preset connectivity conditions; If the connectivity conditions are met, it is determined to be in network mode, the first coordinate data is obtained, and the first coordinate data is sent to the back-end management platform through the broadband communication link so that the back-end management platform can guide the drone to fly based on the first coordinate data; If the connectivity conditions are not met, the system is determined to be in a no-network mode. The second coordinate data is then acquired and sent to the backend management platform via a narrowband satellite communication link in the form of a short message, so that the backend management platform can guide the drone's flight based on the second coordinate data.

2. The UAV cooperative guidance method based on dual-mode communication according to claim 1, characterized in that, The navigation and stopping device includes an RTK positioning module. The step of acquiring second coordinate data and sending the second coordinate data to the backend management platform in the form of short messages via a narrowband satellite communication link includes: Within a preset fitting time window, the target's three-dimensional coordinate time series in the current area is acquired at high frequency based on the RTK positioning module. The three-dimensional coordinate time series of the target is fitted using Chebyshev polynomials. The coefficients of the zero-degree terms in the fitting results are extracted as the reference coordinates representing the absolute position, and the coefficients of the non-zero-degree terms in the fitting results are extracted as the trajectory coefficients representing the relative fluctuations. The reference coordinates are encoded using unsigned integers and a dynamic bit-width allocation algorithm is performed on the trajectory coefficients. The encoded reference coordinates and the trajectory coefficients quantized by the dynamic bit-width allocation algorithm are used together as the second coordinate data. The second coordinate data is combined with the reference timestamp and sampling period features corresponding to the fitting time window to encapsulate and generate a custom short message that meets the single transmission payload limit; Determine whether the current conditions for satellite launch are met based on the locally configured communication frequency smoother; If the conditions are met, the custom short message will be sent to the backend management platform via the narrowband satellite communication link.

3. The UAV cooperative guidance method based on dual-mode communication according to claim 2, characterized in that, The steps of performing polynomial fitting on the target's three-dimensional coordinate time series using Chebyshev polynomials, extracting the zero-degree coefficients from the fitting results as reference coordinates representing the absolute position, and extracting the non-zero-degree coefficients from the fitting results as trajectory coefficients representing relative fluctuations, include: Obtain each physical sampling time corresponding to the fitting time window, and map each physical sampling time to an orthogonal domain normalized variable; Based on the normalized variables, calculate the Chebyshev fitting coefficients of each order; The coefficients corresponding to the zeroth order in the obtained fitting coefficients are defined as the reference coordinates, and the coefficients corresponding to the first to m orders are defined as the trajectory coefficients, where m is the fitting order.

4. The UAV cooperative guidance method based on dual-mode communication according to claim 3, characterized in that, After the step of defining the coefficients corresponding to the first to m terms as the trajectory coefficients, the method further includes: Based on the reference coordinates and the trajectory coefficients, the fitted coordinates corresponding to each of the physical sampling times are reconstructed, and the fitting residual between the fitted coordinates and the target three-dimensional coordinate time series is calculated; Determine whether the fitting residuals meet the preset accuracy conditions. If not, dynamically adjust the fitting order m or the length of the fitting time window based on the fitting residuals.

5. The UAV cooperative guidance method based on dual-mode communication according to claim 2, characterized in that, The step of encoding the reference coordinates using unsigned integers and performing a dynamic bit-width allocation algorithm on the trajectory coefficients, and then using the encoded reference coordinates and the trajectory coefficients quantized by the dynamic bit-width allocation algorithm together as the second coordinate data, includes: Based on a preset scaling factor, the floating-point reference coordinates are converted into unsigned integers of a specified bit width to obtain the absolute reference code value; Obtain the fitting order, and assign corresponding quantization bit widths to each order coefficient in the trajectory coefficients based on the fitting order; Based on the allocated quantization bit width, discretization quantization is performed on the trajectory coefficients of each order to obtain the relative coefficient encoding matrix; The absolute reference code value and the relative coefficient code matrix are concatenated into a bitstream to generate the second coordinate data.

6. The UAV cooperative guidance method based on dual-mode communication according to claim 5, characterized in that, The step of converting the floating-point reference coordinates into unsigned integers of a specified bit width based on a preset scaling factor to obtain the absolute reference code value includes: The values ​​of the reference coordinates on the three spatial axes are extracted respectively, and then multiplied by the scaling factor to map them to the integer domain; The mapped axis values ​​are stored in fixed-point format using 24-bit unsigned integer pairs to obtain the absolute reference code value.

7. The UAV cooperative guidance method based on dual-mode communication according to claim 2, characterized in that, The step of assigning corresponding quantization bit widths to each order coefficient in the trajectory coefficients based on the fitting order includes: Based on the fitting order, obtain the order number k corresponding to each order coefficient in the trajectory coefficients, where 1≤k≤m, and m is the fitting order; According to the order number k in ascending order, the corresponding trajectory coefficients are assigned quantization bit widths in a decreasing trend; Set the upper limit of the total bit width of the relative coefficient encoding matrix corresponding to the trajectory coefficients of each order; If the fitting order reaches a preset order threshold, the load constraint truncation logic is triggered: according to the priority order k from large to small, the bit width of the corresponding trajectory coefficients is compressed in turn, or the items whose order number k exceeds the preset truncation number are directly discarded, until the total bit width of the trajectory coefficients of all orders is less than or equal to the upper limit of the total bit width.

8. The UAV cooperative guidance method based on dual-mode communication according to claim 1, characterized in that, After the step of sending the second coordinate data to the background management platform in the form of a short message via a narrowband satellite communication link, the background management platform is used to receive the second coordinate data sent in the form of the short message, and perform inverse quantization analysis on it to extract the reference coordinates and trajectory coefficients. The back-end management platform is also used to perform continuous interpolation reconstruction based on the reference coordinates and the trajectory coefficients using Chebyshev polynomials to obtain a continuous three-dimensional trajectory curve. The back-end management platform is also used to predict the target heading point based on the continuous three-dimensional trajectory curve, and encapsulate it as a UAV control command to be sent to the UAV's flight control system.

9. A UAV cooperative guidance system based on dual-mode communication, characterized in that, include: Navigation and parking systems, back-end management platforms, and drones; The control device is used to detect the connection status of its own broadband communication link in real time and determine whether the connection status meets the preset connectivity conditions. It is also used to determine that the network is available if the connectivity conditions are met, obtain the first coordinate data, and send the first coordinate data to the background management platform through the broadband communication link; The back-end management platform is used to guide the drone's flight based on the first coordinate data; The navigation device is also used to determine the network-free mode if the connectivity conditions are not met, obtain the second coordinate data, and send the second coordinate data to the background management platform in the form of a short message through a narrowband satellite communication link; The back-end management platform is used to guide the drone's flight based on the second coordinate data.

10. The UAV cooperative guidance system based on dual-mode communication according to claim 9, characterized in that, The guided stop device includes an RTK positioning module, a control module, and a communication frequency smoother; Within a preset fitting time window, the RTK positioning module is used to collect the target three-dimensional coordinate time series of the current area at high frequency, and send the target three-dimensional coordinate time series to the control module; The control module is used to perform polynomial fitting on the target three-dimensional coordinate time series using Chebyshev polynomials, and extract the zero-degree term coefficients in the fitting results as the reference coordinates representing the absolute position, and extract the non-zero-degree term coefficients in the fitting results as the trajectory coefficients representing the relative fluctuations. The control module is also used to encode the reference coordinates using unsigned integers and to perform a dynamic bit width allocation algorithm on the trajectory coefficients, and to use the encoded reference coordinates and the trajectory coefficients quantized by the dynamic bit width allocation algorithm together as the second coordinate data. The control module is also used to combine the second coordinate data with the reference timestamp and sampling period features corresponding to the fitting time window, and encapsulate and generate a custom short message that meets the single transmission payload limit. The communication frequency smoother is used to determine whether the current conditions for satellite launch are met. If the conditions are met, the control module is used to send the custom short message to the backend management platform through the narrowband satellite communication link.

Images