A method for safe flight of unmanned aerial vehicle in interference environment

Abstract

The present invention discloses a method for safe flight of a drone in an interference environment, and belongs to the field of drone control technology. The present invention collects the pitch angle, roll angle, fall speed and fall acceleration of a four-axis drone after the remote control signal is lost; then obtains the height of the four axes based on the pitch angle and roll angle; processes the axis height, fall speed and acceleration through an axis power control model to obtain the control power of the corresponding axis; after the first stage of control of the axis, the control power is compensated based on the height optimization ratio and the speed optimization ratio; the compensated control power is used to perform the second stage of control until the drone enters a stable state. The present invention realizes the stable control of the falling posture of the drone after the remote control signal is lost in an interference environment, thereby ensuring the flight safety of the drone.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle (UAV) control, and in particular to a method for safely flying a UAV in an interference environment. Background Art

[0002] During actual flight, drones often face complex and ever-changing interference environments, such as electromagnetic interference and signal obstruction. These interferences can easily cause drones to lose their remote control signals. Currently, existing technologies for handling drones that lose their remote control signals are relatively simple, mostly relying on pre-set emergency procedures. This makes it difficult to accurately control the drone's attitude and motion in real time, and it's impossible to effectively collect key data when the drone is out of control. This makes it difficult to maintain stable attitude control during a fall, leading to crashes, resulting in economic losses and safety hazards. Summary of the Invention

[0003] In response to the above-mentioned deficiencies in the prior art, the present invention provides a method for safe flight of a drone in an interference environment, which solves the problem in the prior art that after the drone loses the remote control signal, it is unable to stably control its posture during the falling process.

[0004] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is: a method for safe flight of a drone in an interference environment, comprising the following steps:

[0005] S1. Collect the pitch angle and roll angle, as well as the falling speed and falling acceleration of the quadcopter after the remote control signal is lost;

[0006] S2. Obtain the heights of the four axes based on the pitch angle and roll angle;

[0007] S3. Use the axis power control model to process the height, falling speed and falling acceleration of the four axes to obtain the control power of the corresponding axes;

[0008] S4. Performing a first-stage control of the shaft using the shaft control power, compensating the shaft control power based on the height optimization ratio and the speed optimization ratio to obtain a compensated shaft control power;

[0009] S5. Use the compensation control power of the axis to perform the second stage control on the axis until the quadcopter enters a stable state.

[0010] Furthermore, the heights of the four axes are obtained in S2:

[0011] ,

[0012] ,

[0013] ,

[0014] ,

[0015] Where h1 is the height of the first axis, h2 is the height of the second axis, h3 is the height of the third axis, h4 is the height of the fourth axis, r is the length of the arm, and θ is the pitch angle. is the roll angle, and sin is the sine function.

[0016] Furthermore, the axis power control model in S3 includes: 4 axis local power output units, a global power output unit, adder A1, adder A2, adder A3 and adder A4;

[0017] Each axis local power output unit is used to obtain the axis local power according to the difference between the axis height and the maximum height, wherein the maximum height is the maximum value among the heights of the four axes;

[0018] The global power output unit is used to obtain the global power according to the falling velocity and falling acceleration;

[0019] Adder A1 is used to add the global power, the local power of the first axis and the control power of the first axis at the previous moment to obtain the control power of the first axis at the current moment; adder A2 is used to add the global power, the local power of the second axis and the control power of the second axis at the previous moment to obtain the control power of the second axis; adder A3 is used to add the global power, the local power of the third axis and the control power of the third axis at the previous moment to obtain the control power of the third axis; adder A4 is used to add the global power, the local power of the fourth axis and the control power of the fourth axis at the previous moment to obtain the control power of the fourth axis.

[0020] Furthermore, the expressions of the shaft local power output units are: , where P p,t,i is the local power of the i-th axis at the t-th moment, P t-1,i is the control power of the i-th axis at time t-1, h t,max is the maximum height at time t, h t,i is the height of the i-th axis at the t-th time, μ is the denominator coefficient, i is the axis number, and t is the time number.

[0021] Furthermore, the expression of the global power output unit is: , where P o,t is the global power at time t, P t-1,avg is the average value of the control power of the four axes at time t-1, α is the falling acceleration, v is the falling velocity, μ is the denominator coefficient, k a is the falling acceleration weight coefficient, k vis the falling velocity weight coefficient, and t is the time number.

[0022] Furthermore, S4 includes the following sub-steps:

[0023] S41, using the control power of the axis to control the axis in the first stage, recording the control time of the first stage control, and when the control time of the first stage control reaches the time threshold, proceeding to step S42;

[0024] S42. In the first stage control, extract the heights of the four axes at the start and end times;

[0025] S43. At the same time, subtract the height of each axis from the maximum height to obtain a height difference;

[0026] S44, obtaining a height optimization ratio of the corresponding axis according to the height difference between the start time and the end time of each axis;

[0027] S45. In the first stage control, extract the falling speed at the start time and the end time to obtain the speed optimization ratio;

[0028] S46. Obtaining the shaft local compensation power according to the height optimization ratio;

[0029] S47, obtaining a global compensation power according to the speed optimization ratio;

[0030] S48. Add the global compensation power, the axis local compensation power, and the control power of the first stage control output of the corresponding axis at the current moment to obtain the compensation control power of the axis.

[0031] Furthermore, S44 is specifically as follows: for the same axis, subtract the height difference at the start time from the height difference at the end time to obtain an optimized height, and use the ratio of the optimized height to the height difference at the start time as the height optimization ratio of the corresponding axis;

[0032] S45 specifically includes: subtracting the falling speed at the starting moment from the falling speed at the ending moment to obtain the optimized speed, and taking the ratio of the optimized speed to the falling speed at the starting moment as the speed optimization ratio.

[0033] Furthermore, the specific process of S46 includes: when the height optimization ratio is greater than or equal to the height optimization target ratio, the shaft local compensation power is 0; when the height optimization ratio is less than the height optimization target ratio, the shaft local compensation power is calculated according to the difference between the height optimization target ratio and the height optimization ratio: , where P p,H,t,i is the local compensation power of the i-th axis at the t-th moment, P t,i is the control power of the i-th axis at time t, θ H is the local power adjustment step size, ε H,iis the height optimization ratio of the i-th axis, ε H,tar is the target ratio for height optimization, and t is the time number.

[0034] Furthermore, S47 is specifically as follows: when the speed optimization ratio is greater than or equal to the speed optimization target ratio, the global compensation power is 0; when the speed optimization ratio is less than the speed optimization target ratio, the global compensation power is calculated according to the difference between the speed optimization target ratio and the speed optimization ratio: , where P p,v,t is the global compensation power at the tth moment, P t,avg is the average value of the control power of the four axes at time t, θ v is the global power adjustment step size, ε v is the speed optimization ratio, ε v,tar is the target ratio of speed optimization, and t is the time number.

[0035] Furthermore, S5 includes the following sub-steps:

[0036] S51. During the second stage control process, determine whether the falling speed of the quadcopter is less than or equal to 0, and whether the difference between the maximum height and the minimum height of the four axes is less than the height difference threshold. If so, the quadcopter enters a stable state. If not, continue the second stage control until it enters a stable state.

[0037] S52: Maintain the current control power of each axis.

[0038] In summary, the beneficial effects of the present invention are as follows: the present invention first collects the pitch angle, roll angle, fall velocity, and acceleration of the drone, then uses the axis power control model to accurately distribute power, and achieves precise control of each axis of the drone through a compensation mechanism of height optimization ratio and speed optimization ratio. Compared with traditional passive protection solutions, the present invention can quickly respond at the moment of signal loss, minimize the risk of falling, significantly improve the attitude control accuracy of the drone during an uncontrolled fall, quickly stabilize the drone, greatly reduce the risk of crashing due to attitude loss, and improve the safety and reliability of the drone in complex interference environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 The figure is a flow chart of a method for safe flight of a UAV in an interference environment.

[0040] Figure 2 This is a structural diagram of the shaft power control model. DETAILED DESCRIPTION

[0041] The specific embodiments of the present invention are described below to facilitate understanding of the present invention by those skilled in the art. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the concepts of the present invention are protected.

[0042] like Figure 1 As shown, a method for safe flight of a UAV in an interference environment includes the following steps:

[0043] S1. Collect the pitch angle and roll angle, as well as the falling speed and falling acceleration of the quadcopter after the remote control signal is lost;

[0044] S2. Obtain the heights of the four axes based on the pitch angle and roll angle;

[0045] S3. Use the axis power control model to process the height, falling speed and falling acceleration of the four axes to obtain the control power of the corresponding axes;

[0046] S4. Performing a first-stage control of the shaft using the shaft control power, compensating the shaft control power based on the height optimization ratio and the speed optimization ratio to obtain a compensated shaft control power;

[0047] S5. Use the compensation control power of the axis to perform the second stage control on the axis until the quadcopter enters a stable state.

[0048] In this embodiment, the heights of the four axes are obtained in S2:

[0049] ,

[0050] ,

[0051] ,

[0052] ,

[0053] Where h1 is the height of the first axis, h2 is the height of the second axis, h3 is the height of the third axis, h4 is the height of the fourth axis, r is the length of the arm, and θ is the pitch angle. is the roll angle, and sin is the sine function.

[0054] In this embodiment, if Figure 2 As shown, the axis power control model of S3 includes: 4 axis local power output units, global power output unit, adder A1, adder A2, adder A3 and adder A4;

[0055] Each axis local power output unit is used to obtain the axis local power according to the difference between the axis height and the maximum height, wherein the maximum height is the maximum value among the heights of the four axes;

[0056] The global power output unit is used to obtain the global power according to the falling velocity and falling acceleration;

[0057] Adder A1 is used to add the global power, the local power of the first axis and the control power of the first axis at the previous moment to obtain the control power of the first axis at the current moment; adder A2 is used to add the global power, the local power of the second axis and the control power of the second axis at the previous moment to obtain the control power of the second axis; adder A3 is used to add the global power, the local power of the third axis and the control power of the third axis at the previous moment to obtain the control power of the third axis; adder A4 is used to add the global power, the local power of the fourth axis and the control power of the fourth axis at the previous moment to obtain the control power of the fourth axis.

[0058] The first axis local power output unit is used to obtain the axis local power of the first axis according to the difference between the height of the first axis and the maximum height; the second axis local power output unit is used to obtain the axis local power of the second axis according to the difference between the height of the second axis and the maximum height; the third axis local power output unit is used to obtain the axis local power of the third axis according to the difference between the height of the third axis and the maximum height; the fourth axis local power output unit is used to obtain the axis local power of the fourth axis according to the difference between the height of the fourth axis and the maximum height.

[0059] Quadcopter drones rely on the coordinated power of all four axes to achieve attitude and motion control. Each axis handles different forces and attitude adjustments during flight. Local power output units (LPDUs) are installed on each of the four axes to calculate power compensation based on individual axis height differences (caused by pitch and roll angles). For example, if a particular axis is at a relatively low altitude due to the drone's tilt, the LPDU can adjust the power based on the height difference with the highest axis, adapting the power to that axis to its desired position.

[0060] When a drone loses its remote control signal and crashes, it experiences both global motion states, such as overall falling velocity and acceleration, and local height differences due to varying attitudes (pitch and roll) on each axis. The global power output unit (GPO) uses this velocity and acceleration to provide a global power adjustment benchmark to address the drone's overall downward trend. The local power output units (LPOs) address these height differences and address local attitude imbalances.

[0061] In this embodiment, the expressions of the shaft local power output units are: , where P p,t,i is the local power of the i-th axis at the t-th moment, P t-1,iis the control power of the i-th axis at time t-1, h t,max is the maximum height at time t, h t,i is the height of the i-th axis at the t-th time, μ is the denominator coefficient, i is the axis number, and t is the time number.

[0062] The attitude control of a quadcopter drone relies on the power difference of each axis. The height difference of the axis is a direct manifestation of attitude imbalance (for example, the height of the front and rear axes are different when pitching). t,max −h t,i It is the "feedback quantity" that converts the shaft height difference into a basis for power adjustment, allowing the local power of the shaft to change dynamically with the height difference - the greater the height difference, the more adaptable the power correction amplitude.

[0063] The denominator coefficient μ is used to avoid the denominator being 0 and can be a small constant, such as 1 and 0.5.

[0064] In the present invention, t,max −h t,i The bigger the gap, The larger the difference, the greater the local power of the shaft. Therefore, the shaft local power is 0 at the highest shaft, and the shaft local power is the highest at the lowest shaft. During the control process, the difference gradually decreases, and the shaft local power gradually approaches 0.

[0065] In this embodiment, the expression of the global power output unit is: , where P o,t is the global power at time t, P t-1,avg is the average value of the control power of the four axes at time t-1, α is the falling acceleration, v is the falling velocity, μ is the denominator coefficient, k a is the falling acceleration weight coefficient, k v is the falling velocity weight coefficient, and t is the time number.

[0066] When a drone loses its remote control signal and falls, the falling acceleration α reflects the “severity of the change” in the falling trend (such as a sudden rapid acceleration), and the falling velocity v reflects the “speed” of the current fall. The two together characterize the global motion risk. This invention incorporates α and v into the calculation and uses and Adapt and convert the effects of acceleration and velocity respectively, so that the global power can respond to both "falling trend changes" and "current falling rate" at the same time.

[0067] In this embodiment, the falling speed weight coefficient k v and the fall acceleration weight coefficient k a The value range is 0~1.

[0068] In this embodiment, S4 includes the following sub-steps:

[0069] S41, using the control power of the axis to control the axis in the first stage, recording the control time of the first stage control, and when the control time of the first stage control reaches the time threshold, proceeding to step S42;

[0070] S42. In the first stage control, extract the heights of the four axes at the start and end times;

[0071] S43. At the same time, subtract the height of each axis from the maximum height to obtain a height difference;

[0072] S44, obtaining a height optimization ratio of the corresponding axis according to the height difference between the start time and the end time of each axis;

[0073] S45. In the first stage control, extract the falling speed at the start time and the end time to obtain the speed optimization ratio;

[0074] S46. Obtaining the shaft local compensation power according to the height optimization ratio;

[0075] S47, obtaining a global compensation power according to the speed optimization ratio;

[0076] S48. Add the global compensation power, the axis local compensation power, and the control power of the first stage control output of the corresponding axis at the current moment to obtain the compensation control power of the axis.

[0077] In this embodiment, the duration threshold is set to 0.3S.

[0078] After the first stage of control, the present invention extracts the height difference between the starting moment and the ending moment. For the same axis, the height difference between the starting moment and the ending moment is used to obtain a height optimization ratio, which is used to evaluate the optimization of the height difference of each axis. Then, according to the falling speed at the starting moment and the ending moment, a speed optimization ratio is obtained to evaluate the speed optimization. Based on the height optimization and speed optimization, compensation power is obtained respectively to accelerate the stabilization of the drone.

[0079] In this embodiment, S44 is specifically as follows: for the same axis, subtract the height difference at the starting moment from the height difference at the ending moment to obtain the optimized height, and use the ratio of the optimized height to the height difference at the starting moment as the height optimization ratio of the corresponding axis.

[0080] The formula for calculating the height optimization ratio is: , where ε H,i is the height optimization ratio of the i-th axis, H s,i is the height difference at the starting time of the i-th axis, H e,i is the height difference at the end time of the i-th axis.

[0081] S45 specifically includes: subtracting the falling speed at the starting moment from the falling speed at the ending moment to obtain the optimized speed, and taking the ratio of the optimized speed to the falling speed at the starting moment as the speed optimization ratio.

[0082] The formula for calculating the speed optimization ratio in S45 is: , where ε v For speed optimization ratio, v s is the falling velocity at the initial moment, v e is the falling speed at the end moment.

[0083] In this embodiment, the specific process of S46 includes: when the height optimization ratio is greater than or equal to the height optimization target ratio, the shaft local compensation power is 0; when the height optimization ratio is less than the height optimization target ratio, the shaft local compensation power is calculated according to the difference between the height optimization target ratio and the height optimization ratio: , where P p,H,t,i is the local compensation power of the i-th axis at the t-th moment, P t,i is the control power of the i-th axis at time t (P t,i Obtained through the first stage), θ H is the local power adjustment step size, ε H,i is the height optimization ratio of the i-th axis, ε H,tar is the target ratio for height optimization, and t is the time number.

[0084] The present invention is guided by the height optimization target ratio, avoids over-correction through threshold judgment, prevents posture oscillation, and uses the difference between the height optimization target ratio and the actual height optimization ratio to calculate the compensation power, so that the compensation force can accurately match the degree of posture imbalance.

[0085] In the first stage control output of the present invention, the local compensation power of each axis is obtained according to the height optimization ratio, and the global compensation power is obtained according to the speed optimization ratio. The local compensation power of each axis, the control power of the first stage control output of the corresponding axis at the current moment, and the global compensation power are added together to obtain the compensation control power of the corresponding axis.

[0086] In this embodiment, S47 is specifically as follows: when the speed optimization ratio is greater than or equal to the speed optimization target ratio, the global compensation power is 0; when the speed optimization ratio is less than the speed optimization target ratio, the global compensation power is calculated based on the difference between the speed optimization target ratio and the speed optimization ratio: , where P p,v,t is the global compensation power at the tth moment, P t,avg is the average value of the control power of the four axes at time t, θ v is the global power adjustment step size, ε v is the speed optimization ratio, εv,tar is the target ratio of speed optimization, and t is the time number.

[0087] The present invention uses the speed optimization target ratio as a reference and, through threshold judgment, does not perform additional compensation when the speed optimization ratio meets the standard, thereby avoiding power redundancy. When the standard is not met, the difference between the speed optimization target ratio and the actual ratio is used to calculate the global compensation power, so that the compensation strength is adapted to the degree of speed out of control, and the average value of the control power of the four axes is associated, so that the compensation is based on the overall power level and the coordination of power output is guaranteed.

[0088] In this embodiment, small UAVs (such as aerial photography drones, weighing less than 2kg) have small inertia, θ v and θ H Set to 0.1-0.3, small industrial drones (such as plant protection drones, weighing 2-5kg), θ v and θ H Set to 0.3-0.5. Power adjustment step θ H and θ v It can be adjusted according to needs and is not limited to the settings of this embodiment.

[0089] In this embodiment, the height optimization target ratio and the speed optimization target ratio can be set to 0.5. When the ratio reaches 0.5, the height and speed have been optimized by half.

[0090] In this embodiment, S5 includes the following sub-steps:

[0091] S51. During the second stage control process, determine whether the falling speed of the quadcopter is less than or equal to 0, and whether the difference between the maximum height and the minimum height of the four axes is less than the height difference threshold. If so, the quadcopter enters a stable state. If not, continue the second stage control until it enters a stable state.

[0092] S52: Maintain the current control power of each axis.

[0093] After entering the stable state, the control power of each axis is kept unchanged.

[0094] For small consumer drones (such as aerial photography drones with a wheelbase of 20-30 cm), the height difference threshold range is 5-10 cm. For small industrial drones (such as agricultural drones with a wheelbase of 30-50 cm), the height difference threshold range is 10-15 cm.

[0095] In this embodiment, the falling velocity is positive when it is downward and negative when it is upward.

[0096] During the second stage control process, the control power of the first stage control is not directly applied to the four axes, and needs to be compensated in the second stage control process to obtain the compensated control power before it can act on the four axes.

[0097] This method first collects the drone's pitch angle, roll angle, fall velocity, and acceleration. It then uses an axis power control model for precise power distribution. Through a compensation mechanism that optimizes both the height and speed ratios, it achieves precise control of each of the drone's axes. Compared to traditional passive protection solutions, this method can quickly respond to signal loss, minimizing the risk of a fall. It significantly improves the accuracy of the drone's attitude control during an uncontrolled fall, quickly stabilizes the drone, and significantly reduces the risk of a crash due to attitude loss. It also improves the safety and reliability of drones in complex interference environments.

[0098] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for safe flight of a drone in an interference environment, characterized in that: The following steps are involved: S1. Collect the pitch angle and roll angle, as well as the falling speed and falling acceleration of the quadcopter after the remote control signal is lost; S2. Obtain the heights of the four axes based on the pitch angle and roll angle; S3. Use an axis power control model to process the height, fall velocity, and fall acceleration of the four axes to obtain the control power of the corresponding axes, wherein the axis power control model includes: four axis local power output units, a global power output unit, adders A1, A2, A3, and A4; Each of the shaft local power output units is used to obtain the shaft local power according to the difference between the height of the shaft and the maximum height, wherein the maximum height is the maximum value among the heights of the four shafts; , where P p,t,i is the local power of the i-th axis at the t-th moment, P t-1,i is the control power of the i-th axis at time t-1, h t,max is the maximum height at time t, h t,i is the height of the i-th axis at the t-th moment, μ is the denominator coefficient, i is the axis number, and t is the moment number; The global power output unit is used to obtain the global power according to the falling speed and falling acceleration: , where P o,t is the global power at time t, P t-1,avg is the average value of the control power of the four axes at time t-1, α is the falling acceleration, v is the falling velocity, μ is the denominator coefficient, k a is the falling acceleration weight coefficient, k v is the falling velocity weight coefficient; The adder A1 is used to add the global power, the local axis power of the first axis and the control power of the first axis at the previous moment to obtain the control power of the first axis at the current moment; the adder A2 is used to add the global power, the local axis power of the second axis and the control power of the second axis at the previous moment to obtain the control power of the second axis; the adder A3 is used to add the global power, the local axis power of the third axis and the control power of the third axis at the previous moment to obtain the control power of the third axis; the adder A4 is used to add the global power, the local axis power of the fourth axis and the control power of the fourth axis at the previous moment to obtain the control power of the fourth axis; S4. Performing a first-stage control of the shaft using the shaft control power, compensating the shaft control power based on the height optimization ratio and the speed optimization ratio to obtain a compensated shaft control power; S5. Use the compensation control power of the axis to perform the second stage control on the axis until the quadcopter enters a stable state.

2. The method for safe flight of a drone in an interference environment according to claim 1, characterized in that: The heights of the four axes are obtained in S2: ， ， ， ， Where h1 is the height of the first axis, h2 is the height of the second axis, h3 is the height of the third axis, h4 is the height of the fourth axis, r is the length of the arm, and θ is the pitch angle. is the roll angle, and sin is the sine function.

3. The method for safe flight of a drone in an interference environment according to claim 1, characterized in that: The S4 comprises the following sub-steps: S41, using the control power of the axis to control the axis in the first stage, recording the control time of the first stage control, and when the control time of the first stage control reaches the time threshold, proceeding to step S42; S42. In the first stage control, extract the heights of the four axes at the start and end times; S43. At the same time, subtract the height of each axis from the maximum height to obtain a height difference; S44, obtaining a height optimization ratio of the corresponding axis according to the height difference between the start time and the end time of each axis; S45. In the first stage control, extract the falling speed at the start time and the end time to obtain the speed optimization ratio; S46. Obtaining the shaft local compensation power according to the height optimization ratio; S47, obtaining a global compensation power according to the speed optimization ratio; S48. Add the global compensation power, the axis local compensation power, and the control power of the first stage control output of the corresponding axis at the current moment to obtain the compensation control power of the axis.

4. The method for safe flight of a drone in an interference environment according to claim 3, characterized in that: The step S44 specifically includes: for the same axis, subtracting the height difference at the start time from the height difference at the end time to obtain an optimized height, and using the ratio of the optimized height to the height difference at the start time as the height optimization ratio of the corresponding axis; The step S45 specifically includes: subtracting the falling speed at the starting moment from the falling speed at the ending moment to obtain the optimized speed, and taking the ratio of the optimized speed to the falling speed at the starting moment as the speed optimization ratio.

5. The method for safe flight of a drone in an interference environment according to claim 3, characterized in that: The specific process of S46 includes: when the height optimization ratio is greater than or equal to the height optimization target ratio, the shaft local compensation power is 0; when the height optimization ratio is less than the height optimization target ratio, the shaft local compensation power is calculated according to the difference between the height optimization target ratio and the height optimization ratio: , where P p,H,t,i is the local compensation power of the i-th axis at the t-th moment, P t,i is the control power of the i-th axis at time t, θ H is the local power adjustment step size, ε H,i is the height optimization ratio of the i-th axis, ε H,tar is the target ratio for height optimization, and t is the time number.

6. The method for safe flight of a drone in an interference environment according to claim 3, characterized in that: The S47 is specifically as follows: when the speed optimization ratio is greater than or equal to the speed optimization target ratio, the global compensation power is 0; when the speed optimization ratio is less than the speed optimization target ratio, the global compensation power is calculated according to the difference between the speed optimization target ratio and the speed optimization ratio: , where P p,v,t is the global compensation power at the tth moment, P t,avg is the average value of the control power of the four axes at time t, θ v is the global power adjustment step size, ε v is the speed optimization ratio, ε v,tar is the target ratio of speed optimization, and t is the time number.

7. The method for safe flight of a drone in an interference environment according to claim 1, characterized in that: The S5 comprises the following sub-steps: S51. During the second stage control process, determine whether the falling speed of the quadcopter is less than or equal to 0, and whether the difference between the maximum height and the minimum height of the four axes is less than the height difference threshold. If so, the quadcopter enters a stable state. If not, continue the second stage control until it enters a stable state. S52. Maintain the current control power of each axis.

Images