CN-121997769-A - Propelling power consumption modeling and beam forming method for high-altitude platform

CN121997769ACN 121997769 ACN121997769 ACN 121997769ACN-121997769-A

Abstract

The invention provides a propulsion power consumption modeling and beam forming method for a high-altitude platform. The method comprises the steps of firstly constructing a full-system power consumption model covering propulsion, communication and load, and then establishing an accurate propulsion power consumption correction model aiming at pneumatic interference between a boat body and a propeller by utilizing interactive generation type AI intelligent agent to cooperate with CFD numerical analysis, so that deviation of a traditional model is remarkably reduced. On the basis, the invention establishes the beam forming problem of jointly optimizing QoS satisfaction rate and system energy efficiency, and proposes a Q3E algorithm for priority user management. Finally, the non-convex optimization problem is solved rapidly through an unsupervised artificial neural network integrating constraint punishment. The invention effectively solves the problem of power redundancy allocation of HAP in a complex flow field through interdisciplinary modeling and AI reasoning, and remarkably improves the communication energy efficiency while guaranteeing the QoS of users.

Inventors

YANG PENG
Han tengfei
CAO XIANBIN
CHEN YU

Assignees

北京航空航天大学

Dates

Publication Date: 20260508
Application Date: 20260213

Claims (6)

1. The propulsion power consumption modeling and beam forming method for the high-altitude platform is characterized by comprising the following steps of: Step S1, constructing a full-system power consumption model of an altitude platform HAP, covering communication, propulsion, load and maintenance power consumption, and step S2, constructing a propulsion power consumption accurate modeling framework based on generation type artificial intelligence AI energization, and outputting a corrected propeller efficiency model and propulsion power consumption Step S3, based on the propulsion power consumption obtained in step S2 Calculating remaining communication available power budget in combination with HAP total power limit The method comprises the steps of constructing a beam forming optimization problem of joint optimization of QoS satisfaction rate of user service quality and system energy efficiency EE, providing an energy efficiency beam forming algorithm Q3E for enhancing QoS, dividing users into a complete QoS set and a partial QoS set according to user channel state and power requirements, converting an original optimization problem into two cascaded sub-optimization problems, constructing an unsupervised learning architecture based on an artificial neural network ANN, quickly solving the sub-optimization problem, and outputting an optimal power distribution coefficient.
2. The method for modeling and beamforming the propulsion power consumption of the high-altitude platform according to claim 1, wherein the modeling process in step S1 comprises the following steps: Modeling a communication system by providing a HAP with a uniform planar array having respective arrangements in x-axis and y-axis directions And Array element of each antenna, total antenna quantity Power consumption of communication system The method consists of two parts, namely power consumption of a transmitting power dynamic correlation and power consumption of a static circuit, wherein the mathematical expression is as follows: Wherein In order for the power amplifier to consume power, In order for the amplifier to be of low efficiency, For the power consumption of the static circuit, In order to be the total transmit power, For the power consumption of the radio frequency chain, For the local oscillator power consumption, Power consumption is processed for the baseband; propulsion system modeling based on classical momentum theory, propulsion power consumption before AI correction is not introduced Wind resistance to the platform Airspeed of flight The following are related: Wherein For the purpose of the efficiency of the propeller, For motor efficiency, windage The calculation formula is that , In order to achieve a stratospheric air density, As a coefficient of resistance (f) of the material, Is volume; total power consumption synthesis-combining payload power consumption And environmental maintenance power consumption And (3) incorporating to obtain a full-system power consumption equation: 。
3. The method for modeling and beamforming propulsion power consumption of high-altitude platform as claimed in claim 2, wherein the specific content of step S2 comprises constructing a generated AI agent based on a large language model and integrating a retrieval enhancement generation module; the AI intelligent agent receives the geometric shape parameter of HAP and the current flight condition including airspeed And attack angle, automatically writing a control script, calling computational fluid dynamics CFD simulation software, and analyzing the speed field distribution of the tail of the hull by the CFD simulation software; Deriving a correction model, namely performing nonlinear regression analysis on discrete data points output by the CFD by using thinking chain reasoning capability by the AI intelligent agent, aiming at To deduce the efficiency of the propeller taking into account aerodynamic disturbances And airspeed of Is a model of exponential correction: ; Calculating accurate propulsion power consumption, namely coupling the efficiency model with the wind resistance coefficient model of the HAP, and deducing a final accurate propulsion power consumption formula: ; Wherein, the Is the stratosphere air density; is the airspeed of the flight; Is the aerodynamic viscosity coefficient; is HAP volume; the motor efficiency is; the drag correction coefficient is the tail wing drag correction coefficient; is the length of the HAP boat body; is the maximum width of the HAP boat body.
4. The method for modeling and beamforming propulsion power consumption of aerial platform according to claim 3, wherein said step S3 comprises determining a power budget by setting a total energy supply capacity of the HAP to be And (3) accurately propelling power consumption calculated in step S2 And deducting the fixed load and maintenance power consumption from the fixed load to obtain the rest communication available power budget : ; Establishing a channel model by setting the ground existence Users, HAP and the first Channel vector between individual users Modeling as a rice fading channel, comprising a line-of-sight component LoS and a scattering component NLoS; Defining an optimization problem: objective function, joint maximization of system energy efficiency EE and QoS satisfaction rate, EE is defined as the ratio of total transmission rate to total power consumption, qoS satisfaction rate is defined as the actual rate reaching the required rate Is a user proportion of (2); Constraint conditions: The total transmit power does not exceed the budget: Wherein As a function of the power coefficient, Forming a vector for the wave beam; user QoS constraints: users within the set are satisfied for QoS.
5. The method for modeling and beamforming propulsion power consumption of high altitude platform as claimed in claim 4, wherein the step S4 comprises the steps of first receiving the remaining communication available power budget outputted by the step S3 As the upper limit of the total energy constraint of the system, the beam forming optimization problem constructed in the step S3 is received as the target to be solved at the same time, due to the fact that The extrusion value of the pushed power consumption is smaller, constraint conflict is caused by directly solving the optimization problem defined in the step S3, and the Q3E algorithm is adopted to carry out hierarchical processing on the users, wherein the specific implementation process comprises the following steps: calculating a minimum power threshold for each user Calculate that it meets the minimum QoS rate Minimum required signal-to-interference-plus-noise ratio, SINR, to derive minimum transmit power ; Wherein In order for the noise power to be high, For gain, the expression is: ; Wherein, the For the HAP transmit antenna gain, The antenna gain is received for the ground user, In order to transmit the number of antennas, In order to achieve the light velocity, the light beam is, As a function of the carrier frequency, For the height of the HAP flight, The total number of the ground users; full system power check, calculating the sum of the minimum powers of all users ; Case classification and processing: case 1-budget is sufficient, All users can be satisfied, and the optimization goal is to satisfy all On the premise of maximizing the system energy efficiency EE, all users are classified into a set ; Case 2-budget shortage, User discarding is carried out, an algorithm executes a priority ordering strategy, and all users are subjected to minimum power requirements Ascending order is carried out from low to high; Greedy admission-adding users to a service set in order of rank And accumulating the power demand until the accumulated value exceeds Non-selected users are grouped together ; At this time, the optimization objective is converted into a priority maximization set Base of (2), next in the set Internal maximizing energy efficiency EE.
6. The method for modeling and beamforming propulsion power consumption of an aerial platform according to claim 5, wherein the content of step S5 is: inheritance input and network input design, namely, the high-priority user set screened in the step S4 is collected Determining the service object optimized at the time, and extracting a set UPA array response vector for middle user Remaining communication available power budget determined in step S3 QoS requirements of users As an input feature direction of an ANN; the dynamic network architecture is built, namely a five-layer full-connection feedforward neural network FNN adapting to the number change of users is built; An input layer, wherein the number of the neurons is adapted to the dimension of the input characteristics; setting three hidden layers, wherein the number of neurons is preferably 64, 64 and 32 in sequence, and a ReLU activation function is adopted between the layers to introduce nonlinearity and prevent gradient disappearance; output layer arrangement Individual neurons, corresponding to Power distribution coefficient for individual users The output layer adopts a Sigmoid activation function to forcedly map the output on the output layer Interval, representing normalized power ratio; In order to realize the unsupervised learning, a special loss function is designed, constraint conditions are converted into penalty items, and optimization target items are defined firstly System energy efficiency: ; Wherein the method comprises the steps of , The total communication power consumption at the current moment; based on this, a loss function is constructed : ; Wherein, the And To penalize the coefficients, qoS violations are penalized by the ReLU function, and network outputs are forced to strictly meet the total power constraint by the logarithmic barrier function The non-supervision learning without label data is realized; the first term is maximizing energy efficiency, converting negative sign into minimized loss; second term QoS penalty term, user rate Is lower than the requirement Generating forward loss, otherwise, 0; Third, logarithmic barrier term, ensuring that total power is strictly less than Otherwise the loss tends to infinity; Model reasoning, namely deploying a trained ANN model on an HAP (advanced technology attachment) airborne processor, inputting the current channel state and power budget in real time, and directly outputting an optimal power distribution coefficient meeting constraints after forward propagation of a network 。

Description

Propelling power consumption modeling and beam forming method for high-altitude platform Technical Field The invention relates to the technical field of wireless communication and aircraft control intersection, in particular to a propulsion power consumption modeling and beam forming method for a high-altitude platform. Background The High Altitude Platform (HAP) is used as an important component of the 6G space-earth integrated network and is deployed on the stratosphere, and has the advantages of wide coverage, low delay and the like. However, HAPs have limited energy sources (relying primarily on solar energy and batteries) and their propulsion systems consume large amounts of power to resist wind field fluctuations. Traditional researches often neglect the influence of the aerodynamic interference of the boat body on the propulsion power consumption, so that the estimated propulsion power is deviated, and the power budget of the communication load is extruded. Therefore, how to accurately model the propulsion power consumption and optimize the communication power distribution under the constraint is a key to improve the service life and the service quality of the HAP. Disclosure of Invention (One) solving the technical problems The invention aims to overcome the defects of the prior art, provides a propulsion power consumption modeling and beam forming method for a high-altitude platform, and solves the problems in the prior art. (II) technical scheme The invention adopts the following technical scheme for realizing the purposes: A propulsion power consumption modeling and beam forming method of a high-altitude platform comprises the following steps: Step S1, constructing a full-system power consumption model of an altitude platform HAP, covering communication, propulsion, load and maintenance power consumption, and step S2, constructing a propulsion power consumption accurate modeling framework based on generation type artificial intelligence AI energization, and outputting a corrected propeller efficiency model and propulsion power consumption Step S3, based on the propulsion power consumption obtained in step S2Calculating remaining communication available power budget in combination with HAP total power limitThe method comprises the steps of constructing a beam forming optimization problem of joint optimization of QoS satisfaction rate of user service quality and system energy efficiency EE, providing an energy efficiency beam forming algorithm Q3E for enhancing QoS, dividing users into a complete QoS set and a partial QoS set according to user channel state and power requirements, converting an original optimization problem into two cascaded sub-optimization problems, constructing an unsupervised learning architecture based on an artificial neural network ANN, quickly solving the sub-optimization problem, and outputting an optimal power distribution coefficient. Further, the modeling process in the step S1 includes the following: Modeling a communication system by providing a HAP with a uniform planar array having respective arrangements in x-axis and y-axis directions AndArray element of each antenna, total antenna quantityPower consumption of communication systemThe method consists of two parts, namely power consumption of a transmitting power dynamic correlation and power consumption of a static circuit, wherein the mathematical expression is as follows: Wherein In order for the power amplifier to consume power,In order for the amplifier to be of low efficiency,For the power consumption of the static circuit,In order to be the total transmit power,For the power consumption of the radio frequency chain,For the local oscillator power consumption,Power consumption is processed for the baseband; propulsion system modeling based on classical momentum theory, propulsion power consumption before AI correction is not introduced Wind resistance to the platformAirspeed of flightThe following are related: Wherein For the purpose of the efficiency of the propeller,For motor efficiency, windageThe calculation formula is that,In order to achieve a stratospheric air density,As a coefficient of resistance (f) of the material,Is volume; total power consumption synthesis-combining payload power consumption And environmental maintenance power consumptionAnd (3) incorporating to obtain a full-system power consumption equation:。 further, the specific content of the step S2 comprises the steps of constructing a generation type AI intelligent agent, wherein the AI intelligent agent is constructed based on a large language model and integrates a retrieval enhancement generation module; the AI intelligent agent receives the geometric shape parameter of HAP and the current flight condition including airspeed And attack angle, automatically writing a control script, calling computational fluid dynamics CFD simulation software, and analyzing the speed field distribution of the tail of the hull by the CFD simulation softwar