Energy management method for energy power system of low-altitude new energy unmanned aerial vehicle

Abstract

The invention discloses an energy management method for an energy power system of a low-altitude new energy unmanned aerial vehicle. The method comprises the steps of obtaining a current unmanned aerial vehicle state; predicting power requirements of future N steps and solving an optimal power distribution sequence of each subsystem; distributing power to each subsystem and a power battery; each subsystem adjusts the output power according to the adaptive response rate and performs smooth transition control; obtaining the actual output power of each subsystem and the power battery, and updating the SOC of the power battery; updating the efficiency model; and dynamically adjusting parameters of the PID controller. The system has the advantages that each subsystem is ensured to operate at the optimal efficiency point, the utilization rate of hydrogen is greatly improved, and the endurance time of the unmanned aerial vehicle is effectively prolonged; the response rate is adjusted according to the system state to avoid excessive response or insufficient response; an S-shaped curve is introduced, so that power transition is smoother; updating an efficiency model online to adapt to system aging and environment change; and a robust PID controller is adopted, so that external interference and parameter uncertainty are effectively resisted.

Description

Technical Field

[0001] This invention relates to the field of energy management technology, and in particular to an energy management method for a low-altitude new energy unmanned aerial vehicle (UAV) power system. Background Technology

[0002] Currently, the power systems of fixed-wing UAVs are mainly divided into two categories: one is the pure electric system with lithium batteries as the core, which has low energy density and short endurance, making it difficult to meet the requirements of long-endurance missions; the other is the traditional fuel power system, which has a strong endurance, but has problems such as high noise, emissions pollution, and high maintenance costs, making it unsuitable for operation scenarios with high environmental protection requirements or sensitive areas.

[0003] Hydrogen fuel cell systems, as a highly efficient and clean new energy technology, offer advantages such as high energy density, zero emissions, and long endurance, making them an ideal choice for long-endurance power for fixed-wing UAVs. However, fixed-wing UAVs experience drastic power demand fluctuations during different flight phases, including takeoff (when power demand surges dramatically, typically 2-3 times that of cruise), cruise (when power is stable), and landing (when power decreases sharply). Traditional hydrogen fuel cell systems employ a fixed power allocation strategy, which struggles to adapt quickly to sudden power changes, leading to power lag (e.g., insufficient power during takeoff causing climb failure) or energy waste (e.g., maintaining high power output after landing). Furthermore, fixed-wing UAVs often utilize multiple fuel cell subsystems operating in parallel. Existing technologies typically employ average allocation or simple priority strategies for subsystem power allocation, failing to consider the efficiency differences and dynamic characteristics of each subsystem. This can easily result in some subsystems operating in inefficient ranges for extended periods, and even cause problems such as localized overheating and lifespan degradation due to uneven load distribution. Summary of the Invention

[0004] This invention primarily addresses the problems of insufficient dynamic power response and poor multi-subsystem collaborative control in the power systems of low-altitude new energy unmanned aerial vehicles (UAVs). It provides an energy management method for the power systems of low-altitude new energy UAVs, which ensures that each subsystem operates at its optimal efficiency point, greatly improving the utilization rate of hydrogen and thus effectively extending the UAV's endurance. Furthermore, it has the ability to quickly respond to variable load demands during flight, adapting to rapid changes in power requirements during different flight phases, such as takeoff, cruise, and landing, ensuring stable flight of the UAV.

[0005] The technical solution adopted by this invention to solve its technical problem is an energy management method for a low-altitude new energy unmanned aerial vehicle (UAV) power system. The fixed-wing UAV hydrogen fuel cell system includes a power battery and a hydrogen fuel cell. The hydrogen fuel cell includes several subsystems, each of which is independently powered, including:

[0006] S1: Get the current drone status;

[0007] S2: Establish a model predictive control framework, use the model predictive control framework to predict the power demand for the next N steps based on the current UAV state and solve the optimal power allocation sequence for each subsystem;

[0008] S3: Determine whether the total power of each subsystem after being allocated according to the optimal power allocation sequence meets the predicted power demand; if it does, proceed to step S5; if it does not, sort the subsystems according to their efficiency and perform a secondary power allocation on the high-efficiency subsystems.

[0009] S4: Determine whether the total power of each subsystem after the secondary power distribution meets the predicted power demand. If it does, proceed to step S5. If it does not, the power battery will supplement the power gap and proceed to step S5.

[0010] S5: Each subsystem adjusts its output power according to the adaptive response rate and performs smooth transition control;

[0011] S6: Obtain the actual output power of each subsystem and power battery and update the power battery SOC;

[0012] S7: Update the efficiency model;

[0013] S8: Dynamically adjust the PID controller parameters based on the deviation between the actual output power and the predicted power of the hydrogen fuel cell;

[0014] S9: Determine whether the fixed-wing UAV has ended its flight; otherwise, return to step S1.

[0015] As a preferred embodiment of the above scheme, the current UAV status in step S1 includes the power battery SOC, the output power and efficiency model of each subsystem, and the hydrogen cylinder pressure.

[0016] As a preferred embodiment of the above scheme, in step S2, the state equation of the model predictive control framework is:

[0017] x(k+1)=Ax(k)+Bu(k)+ω(k)

[0018] Where x is the state variable x = [SOC, P1, P2, ..., P i ] T u is the control input, u = [u1, u2, ..., u i ] T , where i is the number of subsystems in the hydrogen fuel cell, and P i The real-time output power of subsystem i, u i ω(k) represents the power control command for subsystem i, and ω(k) represents the disturbance process.

[0019] The output equation is

[0020] y(k)=Cx(k)+Du(k)+v(k)

[0021] Where y is the output variable y = [P total Efficiency, SOC T P total denoted as the total output power of the hydrogen fuel cell system for the fixed-wing UAV, Efficiency as the operating efficiency of the hydrogen fuel cell, and v(k) as the measurement noise.

[0022] The objective optimization function is

[0023]

[0024] Among them, y ref (k+j) represents the output value at time k+j, and Q and R are weighting matrices;

[0025] The constraints are

[0026]

[0027] Among them, P i,min and P i,max Let Rate be the minimum and maximum power of subsystem i. max This represents the maximum rate of change of the subsystem power.

[0028] As a preferred embodiment of the above scheme, step S3, which involves sorting the high-efficiency subsystems by efficiency and performing secondary power allocation on them, includes the following steps:

[0029] S31: Calculate the efficiency of each subsystem

[0030]

[0031] Where P is the target power, P upper and P lower These are the two known power points on the efficiency curve that are closest to the target power.

[0032] S32: Sort the subsystems in descending order based on their efficiency;

[0033] S33: Perform secondary power allocation to each subsystem sequentially.

[0034] ΔP i =min(Deficit,P max,i ×PressureFactor-P i,opt )

[0035] Where, ΔP iThe additional power allocated to the i-th subsystem, Deficit is the total power deficit, Deficit is updated after the additional power is allocated to each subsystem, and PressureFactor is the pressure correction factor. P represents the current hydrogen cylinder pressure. max P is the maximum pressure of the hydrogen cylinder. i,opt The optimal power of the i-th subsystem obtained in step S2.

[0036] As a preferred embodiment of the above scheme, in step S5, the adaptive response rate is:

[0037]

[0038] Where BaseRate is the base response rate, P demand For the predicted total power demand, P current This represents the current total power.

[0039] As a preferred embodiment of the above scheme, in step S5, the smooth transition control is achieved by introducing an S-shaped curve instead of linear adjustment.

[0040]

[0041] Where k is the steepness parameter of the S-curve, which is adjusted according to the system response characteristics, and t0 is the transition midpoint time.

[0042] As a preferred embodiment of the above scheme, the power battery SOC update formula in step S6 is as follows:

[0043]

[0044] in P i,act Let C be the actual output power of the i-th subsystem. battery This refers to the capacity of the power battery.

[0045] As a preferred embodiment of the above scheme, in step S7, the efficiency model is updated using the recursive least squares method:

[0046] θ(k+1)=θ(k)+K(k+1)[y(k+1)-φ T (k+1)θ(k)]

[0047]

[0048] Where θ is the parameter vector of the efficiency model, φ is the regression vector, K(k+1) is the gain matrix, P(k) is the covariance matrix, λ is the forgetting factor, 0.95≤λ≤0.99, and y(k+1) is the output measurement of the efficiency model.

[0049] As a preferred embodiment of the above scheme, in step S8, the PID controller is a robust PID controller, and the output deviation of the robust PID controller is as follows:

[0050]

[0051] in, K p K i K d These are the proportional coefficient, integral coefficient, and differential coefficient, respectively. The proportional coefficient K... p Adjusted by parameter adaptive law

[0052]

[0053] σ e Let σ be the standard deviation of the error. e0 K is the initial standard deviation, α is the adaptive gain, and K is the standard deviation. p0 This is the initial scaling factor.

[0054] The advantages of this invention are: it uses a model predictive control framework to predict future power and solve for the optimal power allocation sequence; when the optimal power allocation sequence cannot meet the predicted power, it performs secondary allocation according to the efficiency of the subsystems, ensuring that each subsystem operates at its optimal efficiency point, which greatly improves the utilization rate of hydrogen and thus effectively extends the flight time of the UAV; it adjusts the response rate according to the system state to avoid over-response or under-response; when the power battery SOC is low, it adaptively reduces the response rate to protect the power battery; when the demand changes significantly, it adaptively increases the response rate to quickly track the load; it introduces an S-curve instead of linear adjustment, making power transition smoother; it updates the efficiency model online to adapt to system aging and environmental changes; and it uses a robust PID controller to effectively resist external interference and parameter uncertainties. Attached Figure Description

[0055] Figure 1 This is a flowchart illustrating the energy management method for the power system of a low-altitude new energy unmanned aerial vehicle (UAV). Detailed Implementation

[0056] The technical solution of the present invention will be further described below through embodiments and in conjunction with the accompanying drawings.

[0057] Example:

[0058] This embodiment discloses an energy management method for a low-altitude new energy unmanned aerial vehicle (UAV) power system. The hydrogen fuel cell system of the fixed-wing UAV includes a power battery and a hydrogen fuel cell. The hydrogen fuel cell comprises several subsystems, each of which is independently powered. Figure 1 As shown, it includes:

[0059] S1: Obtain the current drone status, including the power battery SOC, the output power and efficiency model of each subsystem, and the hydrogen cylinder pressure.

[0060] S2: Establish a model predictive control framework, and use the model predictive control framework to predict the power demand for the next N steps based on the current state of the UAV and solve the optimal power allocation sequence for each subsystem.

[0061] S3: Determine whether the total power of each subsystem after being allocated according to the optimal power allocation sequence meets the predicted power demand; if it does, proceed to step S5; if it does not, sort the subsystems according to their efficiency and perform a secondary power allocation on the high-efficiency subsystems.

[0062] S4: Determine whether the total power of each subsystem after the secondary power distribution meets the predicted power demand. If it does, proceed to step S5. If it does not, the power battery will supplement the power gap and proceed to step S5.

[0063] S5: Each subsystem adjusts its output power according to the adaptive response rate and performs smooth transition control;

[0064] S6: Obtain the actual output power of each subsystem and power battery and update the power battery SOC;

[0065] S7: Update the efficiency model;

[0066] S8: Dynamically adjust the PID controller parameters based on the deviation between the actual output power and the predicted power of the hydrogen fuel cell;

[0067] S9: Determine whether the fixed-wing UAV has ended its flight; otherwise, return to step S1. Further, in step S2, the state equation of the model predictive control frame is...

[0068] x(k+1)=Ax(k)+Bu(k)+ω(k)

[0069] Where x is the state variable x = [SOC, P1, P2, ..., P i ] T u is the control input, u = [u1, u2, ..., u i ] T Where i represents the number of subsystems in the hydrogen fuel cell, and in this embodiment, the hydrogen fuel cell has 4 subsystems, P i The real-time output power of subsystem i, u iω(k) represents the power control command for subsystem i, and ω(k) represents the disturbance process.

[0070] The output equation is

[0071] y(k)=Cx(k)+Du(k)+v(k)

[0072] Where y is the output variable y = [P total Efficiency, SOC T P total denoted as the total output power of the hydrogen fuel cell system for the fixed-wing UAV, Efficiency as the operating efficiency of the hydrogen fuel cell, and v(k) as the measurement noise.

[0073] The objective optimization function is

[0074]

[0075] Among them, y ref (k+j) represents the output value at time k+j, and Q and R are weighting matrices;

[0076] The constraints are

[0077]

[0078] Among them, P i,min and P i,max Let Rate be the minimum and maximum power of subsystem i. max This represents the maximum rate of change of the subsystem power.

[0079] Furthermore, in step S3, the secondary power allocation to the high-efficiency subsystems based on their efficiency includes the following steps:

[0080] S31: Calculate the efficiency of each subsystem

[0081]

[0082] Where P is the target power, P upper and P lower These are the two known power points on the efficiency curve that are closest to the target power.

[0083] S32: Sort the subsystems in descending order based on their efficiency;

[0084] S33: Perform secondary power allocation to each subsystem sequentially.

[0085] ΔP i =min(Deficit,P max,i ×PressureFactor-P i,opt )

[0086] Where, ΔP i The additional power allocated to the i-th subsystem, Deficit is the total power deficit, Deficit is updated after the additional power is allocated to each subsystem, and PressureFactor is the pressure correction factor. P represents the current hydrogen cylinder pressure. max P is the maximum pressure of the hydrogen cylinder. i,opt The optimal power of the i-th subsystem obtained in step S2.

[0087] Furthermore, in step S5, the adaptive response rate is...

[0088]

[0089] Where BaseRate is the base response rate, P demand For the predicted total power demand, P current This represents the current total power. The adaptive response rate adjusts according to the system state to avoid over-response or under-response. Using the above adaptive response rate formula, the response rate can be reduced to protect the battery when its SOC is low, and increased to quickly track the load when power demand changes significantly.

[0090] Furthermore, in step S5, smooth transition control introduces an S-curve instead of linear adjustment to achieve smoother power transition, reduce system impact, and extend equipment life.

[0091]

[0092] Where k is the steepness parameter of the S-curve, which is adjusted according to the system response characteristics, and t0 is the transition midpoint time.

[0093] Furthermore, in step S6, the power battery SOC update formula is as follows:

[0094]

[0095] in In actual operation, the total power output of the subsystem may exceed the total power required by the fixed-wing UAV. The excess subsystem output power will charge the power battery. Therefore, P battery P can be negative. i,act Let C be the actual output power of the i-th subsystem. battery This refers to the capacity of the power battery.

[0096] Furthermore, in step S7, the efficiency model is updated using the recursive least squares method:

[0097] θ(k+1)=θ(k)+K(k+1)[y(k+1)-φ T (k+1)θ(k)]

[0098]

[0099] Where θ is the parameter vector of the efficiency model, φ is the regression vector, K(k+1) is the gain matrix, P(k) is the covariance matrix, λ is the forgetting factor (0.95≤λ≤0.99), and y(k+1) is the output measurement of the efficiency model. The real-time updated efficiency model can realize the aging and environmental changes of the hydrogen fuel cell system of the fixed-wing UAV.

[0100] Furthermore, in step S8, the PID controller is a robust PID controller. The robust controller achieves accurate power point tracking, effectively resisting external disturbances and avoiding parameter uncertainties. The output deviation of the robust PID controller is as follows:

[0101]

[0102] in, K p K i K d These are the proportional coefficient, integral coefficient, and differential coefficient, respectively. The proportional coefficient K... p Adjusted by parameter adaptive law

[0103]

[0104] σ e Let σ be the standard deviation of the error. e0 K is the initial standard deviation, α is the adaptive gain, and K is the standard deviation. p0 This is the initial scaling factor.

[0105] The specific embodiments described herein are merely illustrative of the spirit of the invention. Those skilled in the art to which this invention pertains may make various modifications or additions to the described specific embodiments or use similar methods to substitute them, without departing from the spirit of the invention or exceeding the scope defined by the appended claims.

Claims

1. An energy management method for a low-altitude new energy unmanned aerial vehicle (UAV) power system, wherein the power system includes a power battery and a hydrogen fuel cell, and the hydrogen fuel cell includes several subsystems, each subsystem being powered independently, characterized in that: include: S1: Get the current drone status; S2: Establish a model predictive control framework, use the model predictive control framework to predict the power demand for the next N steps based on the current UAV state and solve the optimal power allocation sequence for each subsystem; S3: Determine whether the total power of each subsystem after being allocated according to the optimal power allocation sequence meets the predicted power demand; if it does, proceed to step S5; if it does not, sort the subsystems according to their efficiency and perform a secondary power allocation on the high-efficiency subsystems. S4: Determine whether the total power of each subsystem after the secondary power distribution meets the predicted power demand. If it does, proceed to step S5. If it does not, the power battery will supplement the power gap and proceed to step S5. S5: Each subsystem adjusts its output power according to the adaptive response rate and performs smooth transition control; S6: Obtain the actual output power of each subsystem and power battery and update the power battery SOC; S7: Update the efficiency model; S8: Dynamically adjust the PID controller parameters based on the deviation between the actual output power and the predicted power of the hydrogen fuel cell; S9: Determine whether the fixed-wing UAV has ended its flight; otherwise, return to step S1.

2. The energy management method for the low-altitude new energy UAV power system according to claim 1, characterized in that: The current drone status in step S1 includes the power battery SOC, the output power and efficiency model of each subsystem, and the hydrogen cylinder pressure.

3. The energy management method for the low-altitude new energy UAV power system according to claim 1, characterized in that: In step S2, the state equation of the model predictive control framework is: x(k+1)=Ax(k)+Bu(k)+ω(k) Where x is the state variable x = [SOC, P1, P2, ..., P i ] T u is the control input, u = [u1, u2, ..., u i ] T , where i is the number of subsystems in the hydrogen fuel cell, and P i The real-time output power of subsystem i, u i ω(k) represents the power control command for subsystem i, and ω(k) represents the disturbance process. The output equation is y(k)=Cx(k)+Du(k)+v(k) Where y is the output variable y = [P total Efficiency, SOC T P total denoted as the total output power of the hydrogen fuel cell system for the fixed-wing UAV, Efficiency as the operating efficiency of the hydrogen fuel cell, and v(k) as the measurement noise. The objective optimization function is Among them, y ref (k+j) represents the output value at time k+j, and Q and R are weighting matrices; The constraints are Among them, P i，min and P i,max Let Rate be the minimum and maximum power of subsystem i. max This represents the maximum rate of change of the subsystem power.

4. The energy management method for the low-altitude new energy UAV power system according to claim 1, characterized in that: In step S3, the secondary power allocation to the high-efficiency subsystems based on their efficiency includes the following steps: S31: Calculate the efficiency of each subsystem Where P is the target power, P upper and P lower These are the two known power points on the efficiency curve that are closest to the target power. S32: Sort the subsystems in descending order based on their efficiency; S33: Perform secondary power allocation to each subsystem sequentially. ΔP i = min(Deficit,P) max,i ×PressureFactor-P i,opt ) Where, ΔP i The additional power allocated to the i-th subsystem, Deficit is the total power deficit, Deficit is updated after the additional power is allocated to each subsystem, and PressureFactor is the pressure correction factor. P represents the current hydrogen cylinder pressure. max P is the maximum pressure of the hydrogen cylinder. i,opt The optimal power of the i-th subsystem obtained in step S2.

5. The energy management method for the low-altitude new energy UAV power system according to claim 1, characterized in that: In step S5, the adaptive response rate is: Where BaseRate is the base response rate, P demand For the predicted total power demand, P current This represents the current total power.

6. The energy management method for the low-altitude new energy UAV power system according to claim 1, characterized in that: In step S5, smooth over-control involves introducing an S-curve instead of linear adjustment. Where k is the steepness parameter of the S-curve, which is adjusted according to the system response characteristics, and t0 is the transition midpoint time.

7. The energy management method for the low-altitude new energy UAV power system according to claim 1, characterized in that: In step S6, the power battery SOC update formula is as follows: in P i,act Let C be the actual output power of the i-th subsystem. battery This refers to the capacity of the power battery.

8. The energy management method for the low-altitude new energy UAV power system according to claim 1, characterized in that: In step S7, the efficiency model is updated using the recursive least squares method: θ(k+1)=θ(k)+K(k+1)[y(k+1)-φ T (k+1)θ(k)] Where θ is the parameter vector of the efficiency model, φ is the regression vector, K(k+1) is the gain matrix, P(k) is the covariance matrix, λ is the forgetting factor, 0.95≤λ≤0.99, and y(k+1) is the output measurement of the efficiency model.

9. The energy management method for the low-altitude new energy UAV power system according to claim 1, characterized in that: In step S8, the PID controller is a robust PID controller, and the output deviation of the robust PID controller is as follows: in, K p K i K d These are the proportional coefficient, integral coefficient, and differential coefficient, respectively. The proportional coefficient K... p Adjusted by parameter adaptive law σ e Let σ be the standard deviation of the error. e0 K is the initial standard deviation, α is the adaptive gain, and K is the standard deviation. p0 This is the initial scaling factor.

Images