Method for generating power amplifier model based on genetic algorithm

Abstract

The invention relates to a method for generating a power amplifier model based on a genetic algorithm. The method comprises the following steps: collecting a plurality of corresponding instantaneous input voltages when a power amplifier operates in a saturation region; inputting the plurality of instantaneous input voltages into a memory-free polynomial model, establishing a nonlinear relationship between the plurality of instantaneous input voltages and the plurality of output voltages by using a first coefficient and a second coefficient, and obtaining a corresponding mathematical model when the power amplifier operates in a saturation region, the first coefficient is obtained through genetic algorithm optimization iteration, and the second coefficient is obtained through calculation by using the first coefficient. According to the method, the model fitting precision of the power amplifier can be improved, the complexity of model coefficient solving is effectively reduced, the model solving time is shortened, the model solving efficiency is improved, and ideal model performance is obtained.

Description

Technical Field

[0001] The invention belongs to the technical field of signal processing, and in particular relates to a method for generating a power amplifier model based on a genetic algorithm. Background Art

[0002] The power amplifier is an important component in wireless communication systems. Its function is to amplify the power of the wireless signal to the target value and then feed it into the antenna. During the operation of the amplifier, the input signal is generally amplified in order to output a larger signal. In actual operation, in order to improve the efficiency of the power amplifier, the power amplifier is often operated in the saturation region. In the saturation region of the power amplifier, the output voltage of the amplifier has reached the maximum value, and at this time the gain of the amplifier has begun to decrease, the output signal is distorted, the waveform is distorted, and the input signal can no longer be further amplified. Therefore, when designing and optimizing the amplifier, it is necessary to simulate the nonlinearity of the power amplifier through modeling and establish the mathematical relationship between the input and output of the power amplifier.

[0003] However, existing power amplifier models have problems such as poor fitting accuracy and low computational efficiency, making it difficult to achieve ideal model performance. Summary of the invention

[0004] In order to solve the above problems existing in the prior art, the present invention provides a method for generating a power amplifier model based on a genetic algorithm. The technical problem to be solved by the present invention is achieved by the following technical solutions:

[0005] The present invention provides a method for generating a power amplifier model based on a genetic algorithm, comprising: collecting multiple instantaneous input voltages corresponding to a power amplifier operating in a saturation region; inputting the multiple instantaneous input voltages into a memoryless polynomial model, using a first coefficient and a second coefficient to establish a nonlinear relationship between the multiple instantaneous input voltages and multiple output voltages, and obtaining a mathematical model corresponding to the power amplifier operating in the saturation region, wherein the first coefficient is obtained by optimizing and iterating through a genetic algorithm, and the second coefficient is obtained by calculating using the first coefficient.

[0006] Compared with the prior art, the beneficial effects of the present invention are as follows: in view of the problems that the existing power amplifier model has poor fitting accuracy, low calculation efficiency, etc., and it is difficult to achieve ideal model performance, the present invention proposes a method for generating a power amplifier model based on a genetic algorithm, wherein the method collects instantaneous input voltages at multiple saturation points and uses a memoryless polynomial model to establish a nonlinear relationship between multiple instantaneous input voltages and multiple output voltages, wherein the first coefficient in the memoryless polynomial model is iteratively optimized using a genetic algorithm, and the first coefficient is associated with the second coefficient, and the expression of the entire model function can be obtained by iteratively obtaining the first coefficient, while improving the fitting accuracy of the power amplifier model, effectively reducing the complexity of solving the model coefficients, shortening the model solving time, improving the model solving efficiency, and obtaining ideal model performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 is a flow chart of a method for generating a power amplifier model based on a genetic algorithm provided by an embodiment of the present invention;

[0008] Figure 2 is a fitness evolution curve diagram of the genetic algorithm provided by an embodiment of the present invention;

[0009] Figure 3 is a simulation characteristic curve of the power amplifier model of the present invention provided by an embodiment of the present invention;

[0010] Figure 4 It is an example diagram comparing characteristic curves of an existing power amplifier model provided by an embodiment of the present invention and a power amplifier model generated based on the technical solution of the present invention. DETAILED DESCRIPTION

[0011] The present invention is further described in detail below with reference to specific embodiments, but the embodiments of the present invention are not limited thereto.

[0012] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0013] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification.

[0014] Although the present invention is described herein in conjunction with various embodiments, in the process of implementing the claimed invention, those skilled in the art may understand and implement other variations of the disclosed embodiments by viewing the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other components or steps, and "one" or "an" does not exclude multiple situations. A single processor or other unit may implement several functions listed in the claims. Certain measures are recorded in different dependent claims, but this does not mean that these measures cannot be combined to produce good results.

[0015] The power amplifier is an important component in wireless communication systems. Its function is to amplify the power of the wireless signal to the target value and then feed it into the antenna. The power amplifier has two significant characteristics: nonlinearity and memory. Nonlinearity is mainly manifested in that the gain value of the power amplifier to the input signal changes as the input signal power increases, rather than being a constant. Specifically, when the power amplifier operates in the saturation region, the output power tends to be stable, but the signal may produce a "clipping" phenomenon, resulting in distortion. The "clipping" phenomenon, also known as clipping distortion, can be understood in a popular way: as the input signal power increases, the power amplifier enters the saturation region. From the waveform point of view, the output waveform is similar to the waveform of the input signal, but due to entering the saturation region, the peak part of the output signal (including the top and bottom of the waveform) is cut, resulting in a significant reduction in its amplitude range and waveform distortion. Therefore, it is necessary to accurately simulate the working behavior of the actual power amplifier through power amplifier modeling in order to carry out subsequent design.

[0016] In view of the problems of poor fitting accuracy and low computational efficiency in existing power amplifier models, it is difficult to achieve ideal model performance. The present invention proposes a method for generating a power amplifier model based on a genetic algorithm. This method can improve the fitting accuracy of the power amplifier model while effectively reducing the complexity of solving the model coefficients, shortening the model solving time and improving the model solving efficiency.

[0017] Now, in conjunction with the accompanying drawings, the method for generating a power amplifier model based on a genetic algorithm proposed by the present invention is described in detail. Figure 1 FIG. 1 is a flow chart of a method for generating a power amplifier model based on a genetic algorithm according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0018] Step 110: Collect multiple instantaneous input voltages corresponding to the power amplifier operating in the saturation region.

[0019] Here, the linear frequency modulation signal is input into the power amplifier working in the saturation region, and multiple instantaneous input voltages corresponding to the linear frequency modulation signal and the output voltage of the power amplifier are obtained. The multiple instantaneous input voltages are recorded as x1(t), x2(t),…, x n (t), and record each instantaneous output voltage as y1(t), y2(t),…y n (t), the mathematical relationship between the two is nonlinear.

[0020] It should be noted that, here, the amplitudes of the multiple instantaneous input voltages are positive values, and the method proposed in the present invention is also applicable to the case where the amplitude of the instantaneous input voltage is negative.

[0021] Step 120: Input multiple instantaneous input voltages into a memoryless polynomial model, and use the first coefficient and the second coefficient to establish a nonlinear relationship between the multiple instantaneous input voltages and the multiple output voltages to obtain a corresponding mathematical model when the power amplifier operates in the saturation region, wherein the first coefficient is obtained by iterative optimization using a genetic algorithm, and the second coefficient is calculated using the first coefficient.

[0022] Here, the memoryless polynomial model, also called the non-memory polynomial model, is a general amplifier model. Using the memoryless polynomial model to fit the nonlinear relationship between multiple instantaneous input voltages and multiple output voltages can be simplified to solving each coefficient in the memoryless polynomial model, so that the obtained coefficient can accurately simulate the nonlinear characteristics of the power amplifier. In addition, here, the first coefficient in the model is optimized by the genetic algorithm, which can avoid the problem that the traditional optimization method is easy to fall into the local optimal solution, improve the efficiency and accuracy of the parameter solution, and ensure that the first coefficient is highly correlated with the actual working conditions.

[0023] Here, the mathematical model corresponding to the power amplifier when operating in the saturation region also includes a third coefficient, and the third coefficient is obtained by calculating the gain of the power amplifier. Specifically, the mathematical model corresponding to the power amplifier when operating in the saturation region can be expressed as:

[0024] y(t)=b1x(t)+b3x 3 (t)+b5x 5 (t);

[0025] Among them, x(t) refers to any instantaneous input voltage among multiple instantaneous input voltages, y(t) refers to an output voltage corresponding to x(t), b1 refers to the third coefficient, b3 refers to the second coefficient, and b5 refers to the first coefficient.

[0026] The calculation formula of the third coefficient can be expressed as:

[0027]

[0028] Wherein, b1 refers to the third coefficient, and G refers to the gain of the power amplifier.

[0029] Here, the second coefficient is calculated by the following steps: obtaining the output voltage at the 1dB compression point of the power amplifier, the impedance matching value of the power amplifier, and the output power at the 1dB compression point of the power amplifier; adding the output power at the 1dB compression point and 1dB to obtain the ideal output power; multiplying and square root calculating the ideal output power and the impedance matching value of the power amplifier in sequence to obtain the output voltage at the 1dB compression point; dividing the output voltage at the 1dB compression point and the third coefficient to obtain the input voltage at the 1dB compression point; and calculating the second coefficient using the output voltage at the 1dB compression point, the input voltage at the 1dB compression point, the first coefficient, and the third coefficient.

[0030] Here, the 1dB compression point refers to when the signal strength reaches a certain level during signal input, causing the circuit to enter the nonlinear working area, and the amplifier gain begins to decrease. When the gain drops by 1dB relative to the linear area, the corresponding input or output power point is called the 1dB compression point. In a power amplifier, the 1dB compression point indicates the power point where the amplifier gain begins to compress and decreases by 1dB. This point usually indicates the maximum sustainable level of the output signal. Exceeding this level will cause signal distortion. Specifically, the input voltage at the 1dB compression point can be expressed as follows:

[0031]

[0032] Among them, V i1 Refers to the input voltage at the 1dB compression point, V o1 refers to the output voltage at the 1dB compression point, b1 refers to the third coefficient, P o11 refers to the ideal output power, P o1 Refers to the output power at the 1dB compression point, R ref Refers to the impedance matching value of the power amplifier.

[0033] The output voltage at the 1dB compression point is V o1The mathematical relationship with the input voltage at the 1dB compression point can be expressed as:

[0034] Here, the second coefficient can be calculated by the following formula:

[0035]

[0036] Among them, b3 refers to the second coefficient, b5 refers to the first coefficient, b1 refers to the third coefficient, V i1 Refers to the input voltage at the 1dB compression point, V o1 Refers to the output voltage at the 1dB compression point.

[0037] That is to say, when the third coefficient b1 is a fixed value, the first coefficient b5 is calculated to obtain the value of the second coefficient b3. Compared with the calculation method that requires solving each coefficient separately, the method provided by the present invention can greatly reduce the calculation complexity. When applied to complex calculation scenarios, the calculation results can be obtained faster and the solution efficiency can be improved.

[0038] The method of obtaining the first coefficient b5 is now described in detail. Here, the step of obtaining the first coefficient by optimizing iteratively through a genetic algorithm includes:

[0039] (1) Generate an i-th initial population; the i-th initial population includes a plurality of population individuals, each of which is a random number randomly generated within a preset range; i is an integer greater than or equal to 1.

[0040] Here, when determining the preset range of the first coefficient b5, the power saturation point of the power amplifier needs to be considered. The power saturation point (P1dB) is an important parameter to describe the linearity of the power amplifier. It indicates the input power level when the output power of the power amplifier reaches 1dB compression. Before P1dB, the linearity of the power amplifier is good; when it exceeds P1dB, distortion and nonlinear power amplification will occur. The input voltage at the power saturation point of the power amplifier is recorded as V is , the output voltage is recorded as V os , the mathematical relationship between the two can be expressed as: Since the signal output power reaches the maximum value at the power saturation point, that is, V is >0, there is a point where the derivative is 0, and the input voltage V at the power saturation point of the power amplifier is and output voltage V os After taking the derivative of the mathematical relationship, we can get: Furthermore, the input voltage V is It can be expressed as: According to the existence conditions of extreme points, and V is >V i1 , the preset range of the first coefficient b5 should satisfy:

[0041] The i-th initial population is randomly generated within the preset range, the population size is set to 100, and each population individual is a random number or an estimated value. It should be understood that the population size can be adjusted according to actual needs.

[0042] (2) Encode each individual in the i-th initial population to obtain multiple one-to-one corresponding encoded individuals to form the i-th encoded population.

[0043] Here, binary coding is used to encode each individual in the i-th initial population, according to the precision of 10 -5 Determine the length of the binary bit string. It should be understood that the present invention does not limit the encoding method and encoding type.

[0044] (3) Calculate the fitness value of each coding individual in the i-th coding population one by one, and obtain multiple first fitness values ​​corresponding to the i-th coding population.

[0045] Here, due to the existence of high-order terms b3 and b5, it is difficult to directly solve the mathematical model corresponding to the power amplifier operating in the saturation region, so the problem is transformed into a minimum root mean square error problem, and the fitness value f(b5) of each coding individual is calculated based on the minimum root mean square error of the model. The model function corresponding to the minimum root mean square error is expressed as: in, Refers to the estimated output voltage at the power saturation point, V os Refers to the output voltage value at the power saturation point.

[0046] Here, the fitness value corresponding to each coding individual in the ith coding population is calculated one by one to obtain multiple first fitness values ​​corresponding to the ith coding population, including: for the pth coding individual in the ith coding population, the numerical value corresponding to the pth coding individual is substituted into the mathematical model as the first coefficient, and an output voltage value corresponding to the pth coding individual is calculated; p is an integer greater than or equal to 1; an output voltage estimate value corresponding to the pth coding individual is obtained; the root mean square error of the output voltage estimate value and the output voltage value corresponding to the pth coding individual is calculated to obtain a fitness value, and the fitness value corresponding to the p+1th coding individual in the ith coding population is continuously calculated until the fitness value corresponding to the last coding individual in the ith coding population is obtained.

[0047] Specifically, the fitness value of each coding individual in the i-th coding population is calculated by the following formula:

[0048]

[0049]

[0050] Among them, f i (r) refers to the fitness value of each coding individual in the i-th coding population. Refers to the output voltage estimate of each coding individual in the i-th coding population, V os Refers to the output voltage value of each coding individual in the i-th coding population, V is Refers to the input voltage at the power saturation point of the power amplifier.

[0051] (4) Based on the roulette wheel selection method, multiple first fitness values ​​greater than a preset value are selected from multiple first fitness values ​​corresponding to the i-th coding population, and multiple coding individuals corresponding to the first fitness values ​​greater than the preset value are obtained to obtain multiple selected individuals to form the i-th selection population.

[0052] (5) Perform crossover and mutation processing on multiple selected individuals in the i-th selected population to obtain the i-th selected population after processing.

[0053] Specifically, a crossover process and a mutation process are performed on a plurality of selected individuals in an i-th selected population to obtain an i-th selected population after the process, including: obtaining a preset crossover probability and a preset mutation probability; performing random processing on each selected individual in the i-th selected population to generate a plurality of first random numbers corresponding to each other; for each first random number, when the first random number is less than the preset crossover probability, partially exchanging the selected individual corresponding to the first random number and a selected individual adjacent to the selected individual corresponding to the first random number to obtain the i-th population after the crossover process; performing random processing on each population individual in the i-th population after the crossover process to generate a plurality of second random numbers, wherein one second random number corresponds to a population individual in the i-th population after the crossover process; for each second random number, when the second random number is less than the preset mutation probability, randomly changing a portion of the content of the population individual corresponding to the second random number to obtain the i-th selected population after the process.

[0054] Exemplarily, the preset crossover probability is 0.8, and the preset mutation probability is 0.1.

[0055] It should be noted that when performing crossover processing on the selected individual corresponding to the first random number and a selected individual adjacent to the selected individual corresponding to the first random number, since the selected individuals are specifically numerical values, the crossover processing refers to exchanging part of the numerical values ​​of the two selected individuals. And, when performing mutation processing on the population individual corresponding to the second random number, it refers to changing part of the numerical values ​​of the population individual corresponding to the second random number.

[0056] (6) Calculate the fitness value of each individual in the i-th selected population after processing to obtain multiple second fitness values, and use the maximum value of the multiple second fitness values ​​and the preset accuracy to make a difference to obtain the difference value corresponding to the i-th selected population after processing.

[0057] It should be noted that the method and formula for calculating the second fitness value are the same as the method and formula for calculating the first fitness value, and will not be described in detail for the sake of brevity.

[0058] (7) Determine whether the difference value corresponding to the i-th selected population after processing is less than the preset error value.

[0059] (8) When the difference value corresponding to the i-th selected population after processing is less than or equal to the preset error value, the population individual corresponding to the maximum value among the multiple second fitness values ​​is used as the first coefficient.

[0060] Specifically, after obtaining the first coefficient b5, the second coefficient b3 is calculated according to the mathematical relationship between the first coefficient b5 and the second coefficient b3, and the first coefficient b5, the second coefficient b3 and the third coefficient b1 are substituted into the mathematical model corresponding to the power amplifier operating in the saturation region. In order to be more in line with the actual situation, the noise signal n(t) is added to the model to obtain the final mathematical model corresponding to the power amplifier operating in the saturation region: y'(t)=y(t)+n(t).

[0061] (9) When the difference value corresponding to the i-th selected population after processing is greater than the preset error value, continue to generate the i+1-th initial population for iteration until the first coefficient is obtained.

[0062] Specifically, when the first coefficient cannot be found in the i-th initial population, the i+1-th initial population is randomly generated within a preset range, the population size is set to 100, and each population individual is a random number or an estimated value. Continue to perform the same operation steps as the i-th initial population on the i+1-th initial population to determine whether the first coefficient exists in the i+1-th initial population.

[0063] In order to verify the calculation accuracy of the genetic algorithm, the above steps are simulated using simulation software. Figure 2 : is a fitness evolution curve diagram of the genetic algorithm provided by the embodiment of the present invention. Figure 2 As shown, as the number of iterations of the genetic algorithm increases, the fitness value gradually approaches 0, which means that the model is constantly approaching the optimal solution of the first coefficient. Based on the iterative optimization of the first coefficient using the genetic algorithm provided by the present invention, the nonlinear behavior of the power amplifier can be simulated with high precision.

[0064] Figure 3It is a simulation characteristic curve of the power amplifier model of the present invention provided by the embodiment of the present invention. Among them, the ideal amplification curve refers to the output trend of the power amplifier under the ideal state, and the actual amplification curve refers to the simulation output of the power amplifier model of the present invention. Figure 3 As shown, the actual amplification curve gradually distorts with the increase of input power and deviates from the ideal amplification curve, which is consistent with the working state of the actual power amplifier in the saturation region. In other words, the power amplifier model provided by the present invention can highly simulate the behavior of the actual power amplifier in the saturation region.

[0065] Figure 4 This is an example diagram comparing the characteristic curves of the existing power amplifier model provided by the embodiment of the present invention and the power amplifier model generated based on the technical solution of the present invention. The existing power amplifier model is a linear logarithmic model. The initial simulation conditions are set as follows: the gain of the power amplifier is 13.5dB, the output power corresponding to the 1dB compression point is 21.5dBm, the output power corresponding to the saturation point is 22dBm, the linear amplification gain of the linear logarithmic model is 13.5dB, and the output power corresponding to the logarithmic amplification region threshold is 22dBm. Figure 4 As shown, the two models are simulated. Compared with the linear logarithmic model, the power amplifier model provided by the present invention (referred to as "the model in this article") can simulate the time domain "clipping" phenomenon of the signal level, which means that the power amplifier model provided by the present invention can truly simulate the distortion phenomenon of the power amplifier in the saturation region, and the simulation authenticity is higher.

[0066] In view of the problems that the existing power amplifier models have poor fitting accuracy and low calculation efficiency, and it is difficult to achieve ideal model performance, the present invention proposes a method for generating a power amplifier model based on a genetic algorithm. The method collects instantaneous input voltages at multiple saturation points and uses a memoryless polynomial model to establish a nonlinear relationship between multiple instantaneous input voltages and multiple output voltages. The genetic algorithm is used to iteratively optimize the first coefficient in the memoryless polynomial model, and the first coefficient and the second coefficient are associated. Only the first coefficient needs to be iteratively obtained to obtain the expression of the entire model function. While improving the fitting accuracy of the power amplifier model, the complexity of solving the model coefficients is effectively reduced, the model solving time is shortened, the model solving efficiency is improved, and the ideal model performance is obtained.

[0067] The above contents are further detailed descriptions of the present invention in combination with specific selected implementation methods, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the scope of protection of the present invention.

Claims

1. A method for generating a power amplifier model based on a genetic algorithm, It is characterized in that include: Collecting multiple instantaneous input voltages corresponding to the power amplifier operating in the saturation region; The multiple instantaneous input voltages are input into a memoryless polynomial model, and a nonlinear relationship between the multiple instantaneous input voltages and multiple output voltages is established using a first coefficient and a second coefficient, so as to obtain a corresponding mathematical model when the power amplifier operates in a saturation region, wherein the first coefficient is obtained by iterative optimization using a genetic algorithm, and the second coefficient is calculated using the first coefficient.

2. The method for generating a power amplifier model based on a genetic algorithm according to claim 1, It is characterized in that The step of obtaining the first coefficient by optimizing iteratively through the genetic algorithm comprises: Generate an i-th initial population; the i-th initial population includes a plurality of population individuals, each population individual is a random number randomly generated within a preset range; i is an integer greater than or equal to 1; Encoding each population individual in the i-th initial population to obtain a plurality of one-to-one corresponding encoded individuals to form an i-th encoded population; Calculating the fitness value of each coding individual in the i-th coding population one by one to obtain a plurality of first fitness values ​​corresponding to the i-th coding population; Based on the roulette wheel selection method, a plurality of first fitness values ​​greater than a preset value are selected from a plurality of first fitness values ​​corresponding to the i-th coding population, and coding individuals corresponding to the plurality of first fitness values ​​greater than the preset value are obtained to obtain a plurality of selected individuals to form the i-th selected population; Performing crossover processing and mutation processing on a plurality of selected individuals in the i-th selected population to obtain an i-th selected population after processing; Calculate the fitness value of each population individual in the processed i-th selected population one by one to obtain multiple second fitness values, and use the maximum value of the multiple second fitness values ​​and the preset accuracy to perform a difference to obtain the difference value corresponding to the processed i-th selected population; Determine whether the difference value corresponding to the i-th selected population after the processing is less than a preset error value; When the difference value corresponding to the i-th selected population after the processing is less than or equal to the preset error value, the population individual corresponding to the maximum value among the multiple second fitness values ​​is used as the first coefficient; When the difference value corresponding to the i-th selected population after the processing is greater than the preset error value, the i+1-th initial population is continuously generated for iteration until the first coefficient is obtained.

3. The method for generating a power amplifier model based on a genetic algorithm according to claim 2, It is characterized in that The step of performing crossover processing and mutation processing on a plurality of selected individuals in the i-th selected population to obtain the i-th selected population after processing includes: Obtaining a preset crossover probability and a preset mutation probability; Performing random processing on each selected individual in the i-th selected population to generate a plurality of first random numbers corresponding to each other; For each first random number, when the first random number is less than the preset crossover probability, a selected individual corresponding to the first random number and a selected individual adjacent to the selected individual corresponding to the first random number are exchanged to obtain an i-th population after the crossover process; Performing random processing on each population individual in the i-th population after the crossover processing to generate a plurality of second random numbers, wherein one second random number corresponds to one population individual in the i-th population after the crossover processing; For each second random number, when the second random number is less than the preset mutation probability, a part of the content of the population individual corresponding to the second random number is randomly changed to obtain the i-th selected population after the processing.

4. The method for generating a power amplifier model based on a genetic algorithm according to claim 1, It is characterized in that The mathematical model corresponding to the power amplifier when operating in the saturation region also includes a third coefficient, and the third coefficient is obtained by calculating the gain of the power amplifier.

5. The method for generating a power amplifier model based on a genetic algorithm according to claim 4, It is characterized in that The second coefficient is calculated by the following steps: Obtaining an output voltage at a 1 dB compression point of the power amplifier, an impedance matching value of the power amplifier, and an output power at a 1 dB compression point of the power amplifier; Adding the output power at the 1dB compression point and 1dB to obtain an ideal output power; Performing multiplication calculation and square root calculation on the ideal output power and the impedance matching value of the power amplifier in sequence to obtain the output voltage at the 1 dB compression point; Dividing the output voltage at the 1 dB compression point by the third coefficient to obtain the input voltage at the 1 dB compression point; The second coefficient is obtained by calculation using the output voltage at the 1 dB compression point, the input voltage at the 1 dB compression point, the first coefficient, and the third coefficient.

6. The method for generating a power amplifier model based on a genetic algorithm according to claim 2, It is characterized in that The step of calculating the fitness value of each coding individual in the i-th coding population in a one-to-one correspondence to obtain a plurality of first fitness values ​​corresponding to the i-th coding population includes: For the pth coding individual in the i-th coding population, a numerical value corresponding to the pth coding individual is substituted into the mathematical model as the first coefficient to calculate an output voltage value corresponding to the pth coding individual; p is an integer greater than or equal to 1; Obtaining an output voltage estimation value corresponding to the pth encoding individual; The root mean square error is calculated for the output voltage estimation value and the output voltage value corresponding to the p-th coding individual, respectively, to obtain a fitness value, and the fitness value corresponding to the p+1-th coding individual in the i-th coding population is continuously calculated until the fitness value corresponding to the last coding individual in the i-th coding population is obtained.

7. The method for generating a power amplifier model based on a genetic algorithm according to claim 5, It is characterized in that The input voltage at the 1dB compression point can be expressed as follows: Among them, V i1 Refers to the input voltage at the 1dB compression point, V o1 refers to the output voltage at the 1dB compression point, b 1 refers to the third coefficient, P o11 refers to the ideal output power, P o1 Refers to the output power at the 1dB compression point, R ref Refers to the impedance matching value of the power amplifier.

8. The method for generating a power amplifier model based on a genetic algorithm according to claim 5, It is characterized in that The second coefficient can be calculated by the following formula: Among them, b 3 refers to the second coefficient, b 5 refers to the first coefficient, b 1 refers to the third coefficient, V i1 Refers to the input voltage at the 1dB compression point, V o1 refers to the output voltage at the 1 dB compression point, and G refers to the gain of the power amplifier.

9. The method for generating a power amplifier model based on a genetic algorithm according to claim 4, It is characterized in that The mathematical model corresponding to the power amplifier when operating in the saturation region can be expressed as: y(t)=b 1 x(t)+b 3 x 3 (t)+b 5 x 5 (t); Wherein, x(t) refers to any instantaneous input voltage among the multiple instantaneous input voltages, y(t) refers to an output voltage corresponding to x(t), and b 1 refers to the third coefficient, b 3 refers to the second coefficient, b 5 refers to the first coefficient.

10. The method for generating a power amplifier model based on a genetic algorithm according to claim 6, It is characterized in that The fitness value of each coding individual in the i-th coding population is calculated by the following formula: Among them, f i (r) refers to the fitness value of each coding individual in the i-th coding population, It refers to the output voltage estimate of each coding individual in the i-th coding population, V os It refers to the output voltage value of each coding individual in the i-th coding population.

Images