CN100409580C - Feed forward amplifier system employing self-generating alignment lists and adaptive controller - Google Patents

Abstract

A method and apparatus for rapid alignment of adaptive feed forward power amplifiers are disclosed. Successful alignment settings are correlated with the operating conditions that affect the gain and phase of an amplifier. These operating conditions may include input power level (228), carrier frequency (230), temperature (236), DC supply voltage, or others. The successful alignment settings along with the corresponding operating conditions are stored in a list (224) that is indexed using multi-dimensional attribute vectors. The elements of the list are generated automatically.

Description

Be used to adopt the system and method for the feed-forward amplifier of spontaneous alignment lists and adaptive controller

Related application data

The sequence number that the application requires on January 23rd, 2003 to submit to is the rights and interests of 60/442099 provisional application, and its disclosure is all incorporated herein by reference.

Background of invention

1. invention field

The present invention relates to RF power amplifier and amplification method.More particularly, the present invention relates to feed-forward amplifier and the method that is used to control feed-forward amplifier.

2. description of the Prior Art and related background information

The RF amplifier is the device of attempting to copy the RF signal of present input and producing the output signal with much higher power level.Be called " gain " of amplifier from the power increase that inputs to output.When being constant on the dynamic range of gain at input signal, it is " linearity " that this amplifier is called.Amplifier is finite capacity aspect the power of carrying, because gain and phase change, and saturated under particularly high-power, this makes all actual amplifiers is non-linear when input power levels changes.The distortion power that is produced is measuring of amplifier nonlinearity degree with the ratio of the signal power of conveying.

In the RF communication system, the maximum of amplifier can allow nonlinearity by government organs, as FCC or ITU appointment.Because amplifier is non-linear when working near saturation condition inherently, the linearity requires often to become the restriction to the rated power conveying capacity.In general, when working near saturation condition because the signal power increment that amplifier is carried on the ratio less than the distortion power increment that is produced, the linearity of amplifier degenerates rapidly.

Use various compensation methodes to reduce the distortion of system's output traditionally, thereby improve the rated power conveying capacity.Method for optimizing is a feedforward compensation.In feedforward RF power amplifier, the use error amplifier amplifies the main amplifier distortion component, and this component is synthetic with main amplifier output out-phase again, to offset the main amplifier distortion component.In general, feedforward compensation provides the power capability of main amplifier and the linearity of error amplifier.

The feedforward linearized gain in parallel RF path and the coupling of phase place of being based on of amplifier, thus eliminate carrier wave (input) signal (loop 1) or eliminate distortion (loop 2).Carrier wave is eliminated and is commonly referred to " loop 1 error ", and this is the estimated value of the distortion in main amplifier path.Distortion is eliminated and is carried out in loop 2, uses loop 1 error to eliminate the distortion of main amplifier.The gain in the respective loops and the coupling of phase place are called " loop alignment control ".When the calibration of loop 2 was correct, the distortion in the output was reduced to minimum, makes whole feedforward system more linear than independent main amplifier.When the calibration of loop 1 was correct, the power by error amplifier (it amplifies loop 1 error) was restricted.In most of the cases, before the error amplifier of enabling loop 2, must finish loop 1 calibration.This guarantees that error amplifier is not overdrived, and overdriving is a kind of situation that will produce undesired distortion or device damage.

Most of terminal uses of feed forward power amplifier have the specification requirement of used time of the self adaptation part alignment loop of limit feedforward compensation.As a result, it is important having good initial calibration when adaptive controller begins (or enough) calibrations of search the best.Some these class specification requirements have the time that is low to moderate 10 seconds.

In addition, the good alignment with loop 1 is important with the power that restriction enters error amplifier.But loop 1 error power is the product of input power and calibrating quality (amount of carrier cancellation) haply.As a result, if input power is low, then the calibrating quality of loop 1 may be moderate.If input power increases suddenly, then loop 1 error power will increase pro rata, the error amplifier of might overdriving.Though loop 1 can this increase of response and regulated its alignment settings automatically, the quick variation of the calibration of loop 1 is preferred, the transient effect of error amplifier because it has reduced to overdrive.

For feedforward linearized existing many methods that have earlier, be recorded in nineteen twenties the earliest.In method early, alignment settings is static, and fixing setting is adopted in gain and phase place, is optimized at the nominal operation condition.Method is afterwards introduced look-up table, is used for compensation temperature and DC power source change.Again afterwards, used adaptive approach, the misalignment rate quilt of its intermediate ring road is at internal measurement and be used for the subsequent calibrations adjusting.Loop 1 error power is used as the tolerance of loop 1 misalignment rate usually.In order to measure loop 2, in loop 1, introduce steering signal usually, as " known distortion ".Measured the size of the misalignment rate of loop 2 in the pilot power that output detected of loop 2.Realized being used to measuring the no guidance method of the misalignment rate of loop 2, they are based on distortion measurement.In this type systematic, second loop convergence has significantly and the correlation of the input signal and the distortion that produces.

In the static state of feed-forward amplifier and self adaptation form, all used look-up table.These look-up tables are commonly used to compensation temperature or DC power source change.But even when look-up table is used in the adaptive feedforward system, it is not adaptive part that these look-up tables also normally are used in the control system, such as front-end voltage-controlled attenuator and phase shifter (they keep the entire gain or the phase place of system).That is to say that loop self adaptation and look-up table do not connect.

As noted above, look-up table uses fixed structure in the past always.Input such as temperature is the index of array.These index are distributed on the segment limit equally spacedly by ascending order, in the corresponding alignment settings of array stored.This structure is well suited for memory chip, because index is equivalent to the address, and alignment settings is equivalent to data.But look-up table needs the element of plenty of time ST Stuffing Table usually based on experimental data (calibration).In addition, the drift that is caused by component ageing can make any look-up table become out-of-date, the essential recalibration.

Another difficulty relevant with look-up table is if there are many conditions of work that influence calibrating quality, then need multi-dimension array, and the management multi-dimension array may be very difficult.Can imagine the number of elements that in equally spaced four-dimensional array, exists.For example, 10 samples of every dimension produce 10000 elements.

A kind of technology of managing a plurality of index dimensions is that the supposition effect is discerptible.Discerptible conditions permit is used independent array to each condition of work, and resultant effect is these adjustment sums separately.(be not resemble Taylor series expansion, wherein specify partial derivative)., this method is only adjusted effectively little (differential) calibration, because ignored any cross-correlation between each dimension.Maximum error can appear on the angle of multi-dimension array.For example, the angle of easily ging wrong in temperature, DC power supply index space is high temperature and low pressure.These positions, angle are tested by exper ienced client usually just, so that determine whether this amplifier meets specification requirement.

Be the selection in sampling interval (separation between the adjacent index) in index space based on the relevant problem of the look-up table of array.In general, the susceptibility of gain and phase settings changes on index space.Must come selective sampling density according to the sensitizing range of index space.All the other zones will be by oversampling.The problem of this oversampling is more serious to multi-dimension array.

Attempt to carry out look-up table self calibration or self-generating always.But fixed array structure is difficult to management.The key issue that runs into is ' renewal fragmentation '.Consider above-mentioned four-dimensional array case.When upgrading look-up table, only there is an element to be changed in 10000.If source of degradation is overall (for example being caused by the element drift), then all 10000 elements are all influenced.But change must be propagated along with each index of visit.Exist adjacent index to have the possibility of big-difference, for no other reason than that one of index is older.

Therefore, have the needs to the system and method that is used for the control of feed forward amplifier system rapid loop alignment at present, this system and method has been avoided the above-mentioned limitation of prior art.

Summary of the invention

In first aspect, the invention provides a kind of feed forward amplifier system, comprising first control loop that is used to receive the input of RF input signal and is coupled to input, described first control loop comprises main amplifier, main amplifier sampling coupler, delay element and eliminates combiner.Feed forward amplifier system also comprises second control loop, and it is coupled to first control loop and comprises first signal path, comprises the error coupler of the secondary signal path of error amplifier and first and second signal paths that are coupled.Output is coupled to error coupler.Feed forward amplifier system also comprises and is used for controlling at least one device of first and second control loops, and its adopts the alignment lists with a plurality of list elements, and each element has alignment settings and characterizes the parameter set of the condition of work of feed forward amplifier system.

In a preferred embodiment of feed forward amplifier system, first control loop also comprises fader and phase regulator, and each alignment settings that device adopted that is used to control can comprise the setting of the first control loop fader and the first control loop phase adjuster settings.Second control loop also can comprise fader and phase regulator, and each alignment settings that device adopted that is used to control can comprise the setting of the second control loop fader and the second control loop phase adjuster settings.The parameter set that characterizes the condition of work of feed forward amplifier system can comprise one or more temperature, DC power supply, input signal power and input signal carrier frequency.Preferably, the parameter set that characterizes the condition of work of feed forward amplifier system is defined as attribute vector, and defines distance between any two attribute vectors.The device that is used to control obtains the distance of the attribute vector of current attribute vector and calculating and list element, and the list element that selection has minimum range is as the alignment settings in the controlled function.The device that is used to control can constantly be measured the misalignment rate of feed forward amplifier system, when measured misalignment rate surpasses predetermined value, retrieves alignment settings from alignment lists.The device that is used to control adopts selected element to set as initial calibration, and adopts iterative control algorithm to calculate new alignment settings from initial setting.The device that is used to control upgrades alignment lists with new alignment settings after finishing iterative computation.Distance between the immediate list element attribute vectors can change in whole alignment lists.

According to a further aspect, the invention provides a kind of adaptive controller, be used for the loop of control amplifier system.Adaptive controller comprises one or more inputs, is used to receive the one or more property parameters corresponding to the work at present condition of amplifier system.Adaptive controller also comprises the one or more processors that are coupled to described one or more inputs, described one or more processor has relevant alignment lists, and, thereby provide the loop of loop adjustment setting with the control amplifier system with alignment list algorithm and controller algorithm programming.Alignment list algorithm produces has the alignment lists that adjusting that controller algorithm calculates is set, and one or more property parameters are associated with each adjusting setting.

In a preferred embodiment of adaptive controller, alignment list algorithm is selected alignment settings from alignment lists, uses when starting or when loop becomes fully imbalance for controller algorithm.Alignment list algorithm is preferably by calculating corresponding to the distance between one or more property parameters of work at present condition and the property parameters relevant with each alignment settings in the tabulation, and select alignment settings corresponding to property parameters with minimum range, select alignment lists to regulate and set, use for controller algorithm.Distance calculation can be come weighting with different weights to different property parameters.Property parameters can comprise one or more in temperature, DC supply voltage, input signal power and the input signal carrier frequency.Between two sets of attribute parameters " n " and " 0 " apart from d _AttrCan be for example by the L of weighting _InfThe L of norm distance metric or weighting ₂The norm distance metric defines.Adaptive controller also can comprise one or more inputs, is used to receive calibration data.The one or more inputs that are used to receive calibration data can for example comprise the steering signal input or be used for the input of loop test data.Regulate to set and to comprise fader and phase adjuster settings.

According to a further aspect, the invention provides a kind of method that is used for the control amplifier system, described amplifier system has the control loop that comprises control loop input, first signal path, secondary signal path and control loop output, and at least one in first and second signal paths comprises amplifier.Described method comprises provides list of alignment settings, and each alignment settings has relevant condition of work.Described method also comprises the work at present condition of detecting amplifier system.Described method also comprises in work at present condition and the list of alignment settings those compared, and select with tabulate in the most similar relevant alignment settings of condition of work.

At a preferred embodiment of the method that is used for the control amplifier system, the related work condition is configured to the multidimensional property vector.In work at present condition and the list of alignment settings those are compared distance between each attribute vector in can comprising the current attribute vector of tolerance and tabulating.The alignment settings of selecting to be associated with the most similar condition of work in the tabulation can comprise the attribute vector of determining with the distance minimum of work at present conditional attribute vector.This method can comprise further that also employing iteration loop controller algorithm calculates new alignment settings, and wherein relevant with the most similar condition of work alignment settings is used as the initial calibration of adaptive loop circuit controller algorithm and sets.This method can comprise further that also the new alignment settings that adopts the adaptive loop circuit controller algorithm to be calculated upgrades alignment lists.The size of alignment lists can be dynamic.And, preferably in whole tabulation, change by the interval of institute's stored adjustment settings of attribute vector distance definition.For example, calibration is set the adjusting that higher density can be set in the most responsive zone of the one or more conditions of work that comprise this attribute vector in tabulation.

According to a further aspect, the invention provides a kind of method of list of alignment settings of the control loop of safeguarding amplifier system, described tabulation comprises a plurality of elements, and each element has alignment settings and corresponding to the parameter set of the condition of work of amplifier system.Described method comprises an element selecting alignment lists and determines to have the element of the alignment lists of the relevant work condition the most similar to selected element.Described method also comprises and determines whether that two elements are enough similar and be considered to redundant.Described method also comprises if two elements of alignment lists are redundant, then deletes element the earliest in these elements.

In a preferred embodiment of the method for the list of alignment settings of the control loop of safeguarding amplifier system, select the element of alignment lists to comprise that before not being subjected in the selective listing tabulates and safeguard the element of handling the earliest.Determine to have in the alignment lists relevant work conditions of elements the most similar and can comprise the distance metric of condition of work parameter value of each all the other elements that are determined to alignment lists and the element that selection has minimum range to selected element.Distance metric can comprise corresponding to the weighted difference between the parameter value of condition of work.Can comprise one or more in temperature, DC power supply, input signal power and the input signal carrier frequency corresponding to the parameter of the condition of work of amplifier system.Determine that element is whether enough similar and be considered to redundant and comprise the distance metric determined between the alignment settings and calibrated distance is compared with redundant distance threshold.Perhaps, determine that element is whether enough similar and be considered to redundant comprising the distance between the condition of work parameter of two elements is compared with expired distance threshold.This method can comprise further that also repeating described tabulation for each element in the alignment lists safeguards processing.

According to a further aspect, the invention provides a kind of method of grading list of alignment settings of the control loop that produces amplifier system.Described tabulation comprises a plurality of elements, and each element has alignment settings and corresponding to the relevant parameter collection of the condition of work of amplifier system, and has and comprise the hierarchy of two-stage at least.Described method comprises an element in the first order of selecting alignment lists.Described method also comprises in the first order of determining alignment lists having the relevant work conditions of elements the most similar to selected element, and the earliest the lower one-level that is downgraded to hierarchical alignment list in two elements.

In a preferred embodiment of the method for the grading list of the alignment settings of the control loop that produces amplifier system, determine to have in the alignment lists distance metric of condition of work that the relevant work conditions of elements the most similar to selected element comprises each all the other elements of the first order that is determined to alignment lists, and select element with minimum range.Described method also can comprise determining whether two elements are redundant, and wherein element is early just just demoted under these elements are redundant situation.Described method can comprise that also each level for grading list repeats the tabulation processing.Described method also can comprise if the lowermost level of handling at hierarchy is safeguarded in described tabulation, then deletion clauses and subclauses early.Degraded element is preferably as the not next association of subset list entry of degraded redundant elements.The element of being demoted and having a subset list preferably merges with the subset list of degraded redundant elements not.

According to a further aspect, the invention provides a kind of method that is used for the control amplifier system, described amplifier system has control loop, comprising control loop input, first signal path, secondary signal path and control loop output, at least one in described first and second signal paths comprises amplifier.Described method comprises providing to have the grading list of the alignment settings of two-stage at least, each alignment settings have a alignment settings in the related work condition and the superlative degree part or all have subset alignment settings in more rudimentary.Described method also comprises the work at present condition of detecting amplifier system, and in the superlative degree of work at present condition and alignment settings grading list those are compared.Described method also is included in and selects the alignment settings relevant with the most similar condition of work in the superlative degree of tabulation.Described method also comprises compares in the subclass of work at present condition and selected highest alignment settings those, and selects to have the alignment settings of the subclass of similar condition of work.Described method also comprises and is chosen in the alignment settings that has the condition of work the most similar to the work at present condition in higher or more rudimentary.

In a preferred embodiment, the described method that is used for the control amplifier system comprises each grade reprocessing to grading list, is empty up to the low subclass of the next one.Highest preferably have more rudimentary thicker alignment settings at interval.For example, any two alignment settings can have certain calibrated distance, and the superlative degree has the calibrated distance between the more rudimentary bigger setting.Calibrated distance can comprise the weighted difference between the adjacent setting.For example, regulate setting and can comprise gain-adjusted and phase adjusted setting, and weighting can comprise alignment sensitivity.

Other features and advantages of the present invention are stated in the following detailed description.

Brief description

Fig. 1 is the block diagram according to feed forward power amplifier of the present invention.

Fig. 2 is the block diagram according to the control system of feed forward amplifier system of the present invention.

Fig. 3 is the flow chart according to processing controls algorithm of the present invention, and the pruning that alignment lists is described is with the deletion out-dated elements.

Fig. 4 is the flow chart according to processing controls algorithm of the present invention, and second kind of pruning that alignment lists is described is with the deletion out-dated elements.

Fig. 5 is the flow chart according to processing controls algorithm of the present invention, and the mutual loop alignment control and treatment that comprises between alignment lists processing and the adaptive controller processing is described.

Fig. 6 is the flow chart according to processing controls algorithm of the present invention, and the generation of hierarchical alignment list structure is described.

Fig. 7 be before with clauses and subclauses degradation, the illustrative of hierarchical alignment list structure.

Fig. 8 be after with clauses and subclauses degradation, the illustrative of hierarchical alignment list structure.

Describe in detail

Fig. 1 and Fig. 2 represent the block diagram according to feedforward compensation power amplifier (PA) system of the preferred embodiments of the present invention.Fig. 1 illustrates basic feed-forward amplifier, and Fig. 2 illustrates control system.

As shown in Figure 1, feed-forward amplifier has the traditional architectures that adopts two control loops.Loop 1 comprises signal input part 103, sampling coupler 106, fader 108, phase regulator 109, steering signal input coupler 112, main amplifier 115, main sampling coupler 118, input test coupler 130, delayer 133, eliminates combiner 136 and loop 1 test coupler 139.Loop 2 comprises main sampling coupler 118, main path delay 121, error coupler 124, carrier cancellation combiner 136, loop 2 faders 141, loop 2 phase regulators 142, error amplifier 145, loop 2 test coupler 148 and output 127.As shown in Figure 2, control system can comprise the realization adaptive controller of following detailed description and the processor 202 of alignment list algorithm.Alignment lists is stored in the suitable memory 224, and organizes and access by the mode of following detailed description.Perhaps, can be adaptive controller and alignment list functions processor separately is provided.Digital to analog converter 204,206,208,210 is converted to analog signal α to loop 1 and loop 2 digital gain adjusting and phase adjusted control signal ₁, φ ₁, α ₂, φ ₂, they are provided for gain and phase regulator 108,109,141 and 142 (shown in Fig. 1).The input test data that processor also receives corresponding to the detection input power via power detector 214 and A/D converter 212 from input test coupler 130, and from loop 1 test coupler 139 via power detector 218 and A/D converter 216 reception loops 1 test data.(optionally) steering signal generator 220 produces steering signal, and this signal offers steering signal input coupler 112 (shown in Fig. 1).Steering signal receiver 222 detects any steering signal of not eliminating of self loop 2 test coupler 148 (shown in Fig. 1), and the steering signal that is detected is offered processor 202 with digital form.Processor also receives the input corresponding to the work at present condition of amplifier system.For example, can be provided for the input 226,228,230 of temperature, DC power supply and input rf signal carrier frequency, they are converted to digital form by analog to digital converter 236,238,240.Other condition of work input 234 can be provided and convert digital form to by A/D converter 242.

The general work principle of feed forward amplifier system is at first described.The adaptive feed forward amplifiers control system is by memory and provide quick convergence attribute with successful alignment settings again.This system can learn such as the condition of work of temperature, DC power supply, input power levels and carrier frequency influence that optimal alignment settings had.As a result, the adaptive feed forward amplifiers control system can respond the variation of these conditions of work quickly than the adaptive controller that works separately.

More particularly, the adaptive controller function of processor will be measured the misalignment rate of feedforward loop circuit and regulate calibrator quantity.Controller is regulated calibrator quantity with iterative manner, the search minimum alignment error.In the uncompleted of short duration period of search, the loop imbalance makes the performance degradation of feed forward power amplifier, and too much instantaneous energy is provided, and system linearity is degenerated.In order to make instantaneous deterioration reduce to minimum, good initial calibration is desirable.In addition, good initial calibration meeting reduces to cause unsettled possibility by the adaptive controller of dispersing.

In order to obtain good initial calibration, feed forward power amplifier control system processor 202 is safeguarded the tabulation 224 of before successful alignment settings.Before adaptive controller was used in request, processor 202 was searched for the alignment settings in the past of using under similar condition of work in alignment lists.Condition of work is represented as the multidimensional property vector.Calculate by more current attribute and storing value with undefined attribute ' distance '.In some cases, will provide enough eliminations from the initial calibration of tabulating, thereby avoid the use of adaptive controller.

Attributive distance also is used for the state variation of predictive feed forward power amplifier system.When detecting the unexpected variation of current attribute vector, the feed forward power amplifier system regulates its task arrangement.(noticing that term ' current ' is used in the entire chapter document to describe ' current time ', should not be interpreted as ' electron stream ').Maintenance task is withdrawed from is beneficial to loop alignment, and loop 1 calibration receives limit priority.

Loop 1 calibration relates to by α ₁And φ ₁The control that the represented fader 108 and the adjusting of phase shifter 109 are set, fader 108 can be to press to become attenuator (VVA) or controlled gain device, and they are used to make ' the power minimum that loop 1 test ε (t) ' locates to detect in Fig. 1.The adaptive control function of loop 1 can adopt conventional art, perhaps can adopt the method for the U.S. patent application No.10/733087 that submitted on December 11st, 2003, and it is open all incorporated herein by reference.For loop 2, calibration control can make in that ' loop 2 tests, it is minimum that intermodulation (IMD) vestige that z (t) ' detects reduces to.Alternatively, before main amplifier 115, inject pilot tone 112 and be used as known distortion, make loop 2 calibration controls easier.For example, the usage that pilot tone is used for the loop control of feed forward power amplifier (PA) is disclosed in the U.S. patent No.5796304 that authorizes people such as Gentzler, disclosing of this patent is incorporated herein by reference.These loop 2 adaptive control functions can adopt conventional art, perhaps, also can adopt the method for the U.S. patent application No.10/733498 that submitted on December 11st, 2003, and disclosing of this patent application is all incorporated herein by reference.In this document, fader and phase shifter values are called as ' alignment settings '.The skew of fader and phase shifter settings and its optimum value is called as ' misalignment rate '.When the skew of setting with the best hour, loop is called as ' calibration '.

Many conditions of work that influence optimal alignment settings are arranged.These comprise environmental condition, use certain conditions and system's certain conditions.For example, it is the related work condition that temperature, input power levels x (t), carrier frequency and DC supply voltage are used great majority, and these are used as input 226,214,230 and 228 and provide, as shown in Figure 2.Time also can be regarded as influencing the parameter of calibration, because element can wear out.Loop 1 alignment settings also can influence optimal loop 2 alignment settings.These parameters all can be measured in the feed forward power amplifier system, and can be monitored by processor 202.In control system of the present invention, relevant measurable parameter is used to constitute attribute vector.Attribute vector when loop alignment and related setting are stored in the alignment lists in the memory 224.Attribute vector is realized by alignment lists with the relevant of alignment settings.Disclosed system handles alignment lists with the adaptive controller processing and combines to allow rapid loop alignment.

Below, with reference to Fig. 1-8 specific embodiment of the present invention is described.

At first, expression attribute vector and alignment settings and it is stored in the preferred embodiment of the mode in the alignment lists are described.The distance metric of the similitude that is used for the metric attribute vector is also described.Also discussed and be used to prune the process of alignment lists to allow to safeguard the list of entries that to manage quantity, the covering that keeps attribute space simultaneously.Then, describe alignment lists and adaptive controller and be used for the usage that regulation loop is set.The self-generating of the element in the alignment lists is also described.Then in conjunction with the alternative alignment list structure of Fig. 6-8 description based on hierarchical list structure.

As mentioned above, adopt sets of attribute parameters or attribute vector to allow the feed forward power amplifier system from past actions, to learn.By relevant with alignment settings in the past attribute vector, it is faster and more healthy and stronger that loop convergence becomes.For given application, the property parameters that influences the feed forward power amplifier system calibration is determined when the defined attribute vector.Temperature, Mean Input Power and centre frequency are important parameters for cellular application usually because they become when all being and the influence gain.Parameter such as modulation format and number of carriers also influences gain; But if these parameters do not change in time, they provide very little value as the part of attribute vector, can get rid of in the dependency vector.

In order to determine the similitude of two attribute vectors, adopt distance metric.Difference between each parameter (difference of two squares or absolute difference) is come weighting according to its susceptibility for the feed forward power amplifier system gain.That is to say that the parameter that the feed forward power amplifier system gain is had considerable influence is by weighting more.These susceptibilitys can utilize part specifications or estimate for the experiment of specific implementation and application requirements.

Specifically, allow the attribute vector about alignment list element ' n ' be defined as

(formula 1) p (n)=[p ₀(n) ... p _m(n)]

P wherein _kIt is the value of property parameters ' k ' (for example temperature).Realize on processor 202 for the ease of distance metric, can use the L after the weighting _InfNorm: that is to say, by d _AttrElement ' n ' and the distance between ' 0 ' that (n, 0) is represented are defined as:

(formula 2) d _Attr(n, 0)=max _k{ w _k| p _k(n)-p _k(0) | }

W wherein _kIt is the weight of parameter ' k '.Also can adopt alternative distance metric, for example L ₂Norm.By d _AttrL after the element ' n ' that (n, 0) is represented and the weighting of the distance between ' 0 ' ₂Norm measure is defined as:

(formula 3) $d_{\mathrm{attr}}(n,0)={\left[\underset{k}{\mathrm{\Σ}}\right\{w_k\·{|p_k\left(n\right)-p_k\left(0\right)|}^2\left\}\right]}^{0.5}$

The similitude of tolerance alignment settings also is valuable.Can reuse L _InfNorm:

(formula 4) d _Align(n, 0)=max _k{ s _k| Δ _k(n, 0) }

Δ wherein ₀(n, 0)=VVA (n)-VVA (0), Δ ₁(n, 0)=PHS (n)-PHS (0), wherein VVA (n) is that n gain calibration is set, PHS (n) is that n phase alignment set s _kIt is corresponding susceptibility.Sensitivity terms can make apart from d through selecting _AlignCorresponding to working as [VVA (0), PHS (0)] elimination residual error when being best alignment.Also can adopt for example L once more ₂The alternative distance metric of norm.d _AlignL ₂Norm measure is:

(formula 5) $d_{\mathrm{align}}(n,0)={\left[\underset{k}{\mathrm{\Σ}}\right\{s_k\·{\left|{\mathrm{\Δ}}_k(n,0)\right|}^2\left\}\right]}^{0.5}$

Alignment list structure can be a dynamic.List of entries and number of entries all can dynamically change.More particularly, with attribute vector, processor 202 is followed the tracks of in the past successful alignment settings.Before realizing the adaptive controller function, the residual error of the current alignment settings of processor 202 tests.If it is fully with regard to eliminating quality, then without any need for action.If it is not fully, then retrieve the alignment settings of listing element that has with the immediate attribute vector of work at present condition.Test the residual error of new settings then.If eliminate still insufficiently, then adaptive controller produces new alignment settings.After given loop has been restrained,, new alignment settings is added in the alignment lists by means of adaptive controller.

The number of elements of restriction in the alignment lists is with limiting processor computing time, and this is important.The simplest method is that number of elements is provided with the upper limit, and if received new alignment settings, just cover element the earliest.A kind of alternative approach is to use pruning.When processor 202 is not busy with the processing priority order, can carry out pruning.When almost completely consistent than the alignment settings person of being adjacent of element early, pruning is usually eliminated redundancy by deletion than unit early and (be that is to say that the neighbor of element ' k ' has minimum d _Attr(n, k), and if d _Align(n, k) little, then be considered as redundancy).As a result, the ratio that is varied in the representation attribute space number of elements that the zone is required and the alignment settings.This method based on tabulation produces the concise representation of attribute-alignment mapping.

Only, represented a kind of realization of alignment lists pruning process stream among Fig. 3 as example.As shown in the figure, when processor 202 was not taken by the task of higher priority, process flowed since 302.Process flows to 304 to select element the earliest from alignment lists.Then, the attribute vector distance metric d that defines as mentioned of processing and utilizing _AttrDetermine that apart from the immediate element of all the other elements of tabulation concrete mode is,, select element 308 then with minimum range at first in 306 distances that calculate all the other elements of tabulation.Then, 310, handle flowmeter and calculate the calibrated distance d of definition as mentioned _AlignTo select immediate element.If alignment settings is identical or enough approaching, then at 312 deletions elements early.Enough near depending on elimination tolerance limit and elimination susceptibility to respective alignment settings (fader or phase shifter).For example, can adopt minimum range d _RedundantIf, and d _AlignBe less than or equal to d _Redundant, then deletion is than element early, and if d _AlignGreater than d _Redundant, then keep two elements.314, process stream checks whether whole tabulation is examined, if do not have, then repeats until all the other elements separate enough (perhaps up to the lower limit that reaches list size) far away with regard to fader and phase shifter tolerance limit.

The second pruning process is shown among Fig. 4.This can be used to delete out-dated elements from alignment lists.For example, the component ageing meeting changes the relation between optimal alignment settings and the given attribute vector.Little inter-element attribute distances has identified out-dated elements potentially.For example, if element has incorrect alignment settings, then residual error will be too big, makes to use the adaptive controller function.Adaptive controller can find new alignment settings.As a result, two different calibrations will be listed certain given attribute vector (or two closely similar vectors).Deletion element has early solved any conflict and has kept tabulation for current.

With reference to this second particular procedure stream pruned that is used for shown in Figure 4, when processor was not carried out the task (comprise first prune handle) of higher priority, second pruned and handles in 402 beginnings.404, process stream is selected in the alignment lists element the earliest, and it is pruned at second as yet and handles.406, for this element the earliest calculate in the tabulation in all the other elements each apart from d _AttrThen 408, handle the minimum range d that determines to have to the earliest element _AttrElement.If this minimum range is less than or equal to predetermined distance d _Outdated,, handle and from alignment lists, delete in two elements than early one then 410.But, if minimum range is greater than d _Outdated, then these two elements are considered to enough differences, and the both is retained in the alignment lists.412, process stream checks whether also have element to check, if then process stream turns back to 404 to check next element the earliest in the tabulation.After all elements all was examined, the second pruning process withdrawed from 414.

At alignment lists is under the situation of sky, can keep the default alignment of each loop to set.And, can keep factory default alignment to set dividually, make it can not pruned.

Then, with reference to Fig. 5, the loop alignment control and treatment is described.Each alignment loop control procedure (being loop 1 control and loop 2 controls) has similar structure, and it is described in Fig. 5.The feed forward power amplifier system preferably is that each loop (above about described loop 1 of Fig. 1 and loop 2) is safeguarded alignment lists separately.

Two parallel procedures 500 and 501 and control interactional process stream between these two processes are used in expression as among Fig. 5, loop alignment algorithm.First process that 501 places indicate produces new alignment settings, handles 500 for initially being used for the adaptive loop circuit controller.This process 501 is used when starting, and operation continuously after this, handles to allow being used for controller when misalignment rate becomes too big.More particularly, first process flows 501 continuous monitoring work at present conditional parameters to obtain current attribute vector, shown in 504.For example, processor can be determined Current Temperatures, DC power, carrier frequency and input power, to determine the current attribute vector of work at present condition.Then, 506, process stream is checked alignment lists, the list element of the attributive distance minimum of searching and current vector.Then in 510 retrieval these list element alignment settings (loop 1 or loop 2 alignment settings).The alignment settings of this retrieval whether be used for upgrading adaptive controller handle 500 calibration can by misalignment rate measure handle 514 and the variation of the element of retrieving control.When starting,, then retrieve alignment lists and set and be used for upgrading these calibrations 512 as long as measured misalignment rate is too big.Initiate adaptive controller then and handle 500, and adaptive controller then 516 for loop 1 or loop 2 utilize iteration control device algorithm computation calibration correction value, as previously described.

For example, this process stream can be controlled by Status Flag.If the element in 510 retrievals changes, then Status Flag is cleared to show system's ' not iteration '.Process stream is also at 514 continuous measurement misalignment rates.If misalignment rate is too big, then check Status Flag.If sign is cleared, then shows the great change of attribute vector, and utilize the new settings of from alignment lists, retrieving to upgrade calibration.After retrieving novel was set, Status Flag was set with beginning iteration self-adapting controller process.If sign is set, then ask adaptive controller process 500, this process is calculated the loop misalignment rate 516 according to discrepancy adjustment.Upgrade calibration 518, the iteration process is calibrated until loop.If misalignment rate measured after finishing the adaptive controller processing is little, then be stored in the alignment lists in 520 alignment settings and current attribute vector.Finish loop alignment routine then.At this moment, the different loop of adjustable or can prune various alignment lists as described above.When system closing or after overtime, in 522 optimal alignment settings that can select session, this can be stored and be used to initiate quick startup.

With reference to Fig. 6-8, utilize the hierarchy of element that the additional optional feature that alignment lists is handled is described.In said method, when the number of elements in the selection tabulation, exist compromise.Allowing the benefit of a great number of elements is the intensive covering of attribute space.But the benefit of oligo-element is to need the less time to determine element with the attributive distance minimum of current vector.Utilize hierarchical alignment list structure to allow intensive covering and search fast.

The use of pruning can be used to produce hierarchical alignment list, shown in Fig. 6-8.In the described in front redundant pruning process, when determining two elements " similar ", element early becomes redundancy and deleted, only has another element to survive.In the stage division of list management, redundant elements is not deleted; But be " demoted " to lower subsets below the survival element.The establishment of lower subsets is a recurrence, and this allows the grade of definition requirement.(in most of the cases, zero or a subclass grade just enough).

The basic pruning process stream that is used for producing hierarchical alignment list is shown among Fig. 6.602, pruning begins when processor is not taken by higher priority task.604, the clauses and subclauses of identification redundancy are right.This process 604 can be according to same distance calculation described in conjunction with Figure 3 (304,306,308 and 310).606, being labeled as redundant entries than clauses and subclauses early, another then is retained as its current grade in hierarchical list structure.608, pruning process stream checks whether tabulation is in the lowest class.If, then in 610 deletion redundant entries.If tabulation then makes redundant entries be downgraded to next lower grade 612 not at the lowest class.Its position in lower grade is marked as the sub-list of entries of the survival clauses and subclauses in the higher level.Then, flow to into next lower grade in 614 pruning processes, and handle 602 in this grade to current clauses and subclauses tabulation beginning.

Redundant elements can have subclass.Such hierarchy is shown among Fig. 7 and Fig. 8.Before redundant elements 710 was by degradation, its lower level subset tabulation 714 merged (in same grade) with the subset list 712 of survival element.Redundant elements 710 is downgraded to level grade 704 then.From viewpoint at the moment, merge the subclass that is in same grade and make subclass become too big; But when initiating the pruning process, the too little meeting of subclass turns back to desirable value.

In grading list, highest ranking 702 preferably has the most rough sampling (thresholding d _AlignMaximum).Low more grade has meticulous more resolution (more little d _Align).By adjusting the thresholding d of each grade _Align, might be adjusted at the number of entries in the given alignment lists of given grade.By increasing d _Align, the number of entries in the tabulation reduces.Hope all tabulations on the various grades of hierarchy have almost equal number of entries.

The loop alignment process that is used for grading list flows generally the process of Fig. 5 stream 501 as described above.But, be limited to a set 506 pairs of search with element of minimum attribute distance; This set is called ' effectively '.When search 506 beginnings, top-level set is effective.To the search of the alignment settings of approaching current attribute vector at first search in the highest ranking tabulation near clauses and subclauses, search for the subset list of these clauses and subclauses then.In case identify the element that has minimum attribute distance in the highest ranking, then retrieve its calibration (as above in conjunction with as described in Fig. 5).But, not to ask adaptive controller as mentioned above, the next lower level subset of element becomes effectively.Search for next lower level subset then searching element with minimum attribute distance, and the alignment settings of retrieving novel.Process recursively repeats, and is empty up to the next one than lower subsets.At this moment, request adaptive controller and adaptive controller use the alignment settings of being retrieved.This process recursively repeats, up to reaching the lowest class.If arbitrary intermediate settings provides enough calibrating qualities, then process was ended before the request adaptive controller.Notice that subset search should comprise father's clauses and subclauses (if two or more grades are perhaps arranged below the superlative degree, then should comprise a plurality of father's clauses and subclauses), because it may be an optimum Match.

Amassing of grade quantity (L) in the complexity of search time and number of entries of each tabulation (N) and the hierarchy is proportional.On the contrary, the L power of exhaustive search and N is proportional, and this is all obviously much higher usually.(each tabulation of this supposition has N clauses and subclauses.Notice that the subset list that grade L has is N a times of grade L-1, this means that grade L has the L power clauses and subclauses of N altogether).

Noting, is expired as the element of pruning the part deletion of process in conjunction with Fig. 4 described second, therefore should not be kept in the set of lower grade.

In sum, disclosed system handles the multidimensional calibrating tabulation with the adaptive controller processing and combines, with the control loop of calibration feed forward amplifier system.These two processes combine in a novel way to improve the dynamic response of system.The tabulation of the multidimensional calibrating that uses in the disclosed method have with based on the different structure of the look-up table of array, thereby avoid the problems referred to above of look-up table.Not to utilize array structure to come storage element, but element is put together as set.Each element has following: (a) corresponding with the condition of work that influences amplifier one group of parameter or attribute; And the optimal alignment settings of (b) under those conditions of work, searching.Form a kind of tolerance, ' distance ' between two elements of its definition, this is based on the difference between the element property.If feedforward system detects the imbalance in the given loop, then measure and work at present condition associated attributes.Then, in the identification alignment lists with the minimum element of the distance of current attribute, and from memory its corresponding alignment settings of retrieval.If new alignment settings is unsuitable, then adaptive controller is activated with the further loop alignment of improving.In case think calibration enough near optimum value, then alignment settings is combined to form the new element in the set with current attribute.Therefore, alignment lists is spontaneous.

In order to limit the computation complexity of search minimum distance element, need the size of constraint element set.In order to discern redundant elements, measure the similarity (carrying out during at one's leisure) of element with attributive distance and alignment separation.If set sizes surpasses preset quantity, then delete similar element centering element the earliest.By the restriction list size, definite time apart from minimum element with current attribute is controlled.If need additional elements for bigger coverage, then might form the hierarchy of list level.Not the similar element of deleting the earliest, but be " demoted " to the following lower level subset of survival element.If the element of degradation comprises its own lower level subset, then with the subclass merging of they and survival element.Establishment than lower subsets is a recurrence, and this allows the grade of definition requirement.(in most of the cases, zero or a subclass grade just enough).

It is that the last efficient of calculating is high that hierarchy allows the search to minimum distance element, because each efficient set is big or small limited.At first, the set of highest ranking is effective.Do not produce enough calibrations if having the element of minimum range in the set of highest ranking, then its subclass (if there is) becomes effectively.Test minimum distance match at calibrating quality from lower level subset.Search subset recursively, up to calibrating quality be enough or next be empty than lower subsets.For back one situation, adaptive controller will be activated to improve calibration.

The work in combination of adaptive feedforward controller and disclosed alignment lists can be learnt disclosed amplifier system from its experience, and improving its performance when having dynamic condition, described dynamic condition comprises such as the input power levels that changes, the carrier frequency of variation (or saltus step), temperature or the DC power supply or the component ageing of variation.Any amount of attribute of system's tolerable (multi-dimensional indexing space) and can obviously not increase complexity.The hierarchical set management allows the element of the arbitrarily big quantity of storage, and the worst condition that can obviously not be increased in when searching the element with minimum attribute distance postpones.

Another advantage of method that is used for the adjusting of first loop (carrier wave elimination) is that the transient power that passes through error amplifier that occurs is reduced to minimum when input power increases suddenly, thereby has reduced the possibility of damaging.This is for using the Amplifier Design particular importance of little error amplifier.For the adjusting of second loop in not having the guiding realization, when the vacation detection was impossible, this method provided stable default value.This is again an advantage for some Amplifier Design.

The additional benefit of disclosed alignment lists is that attribute space can be sampled unevenly.In general, the susceptibility of gain and phase settings changes on attribute space.Disclosed system intentionally produces higher element density in the zone that has than high sensitive certainly.

Alignment lists provided by the invention is useful for dynamic waveform.For example, a kind of application of feed forward power amplifier system is to be used for the frequency hopping beacon.In this application, the alignment settings of each carrier wave can be stored fast frequency-hopped to allow (with respect to typical PA setting-up time).

Consider foregoing, be appreciated that to the invention provides a plurality of favourable features.Alignment lists provides the quickly calibrated of loop with being used in combination of adaptive controller processing in the feed forward power amplifier system.Alignment lists allow system from the experience in past, to learn from natural disposition, reduce adaptive controller required search time.Reduced the transient energy surge, and adaptive controller is when obtaining gaining and becoming healthy and strong more during the good initial estimated value of phase settings.It is that computational efficiency is high that the hierarchy of alignment lists allows the search to minimum distance element, and the wide coverage of attribute space is provided simultaneously.

The present invention is described in conjunction with presently preferred embodiment, still, it will be understood by those of skill in the art that and can carry out various modifications, can't describe one by one too much, still within the scope of the invention simultaneously.Therefore, describe more than that should to be considered as only be illustrative in detail, rather than determinate.

Claims (50)

1. feed forward amplifier system comprises:

Be used to receive the input of RF input signal;

First control loop is coupled to described input and comprises main amplifier, main amplifier sampling coupler, delay element and elimination combiner;

Second control loop is coupled to described first control loop and comprises first signal path, comprises the error coupler of the secondary signal path of error amplifier and described first and second signal paths that are coupled;

Be coupled to the output of described error coupler; And

Be used for controlling at least one device of described first and second control loops, its adopts the alignment lists with a plurality of list elements, and each element has alignment settings and characterizes the parameter set of the condition of work of described feed forward amplifier system.

2. feed forward amplifier system as claimed in claim 1, it is characterized in that, described first control loop also comprises fader and phase regulator, and each described alignment settings comprises that the first control loop fader is set and the first control loop loop phase adjuster is set.

3. feed forward amplifier system as claimed in claim 1, it is characterized in that, described second control loop also comprises fader and phase regulator, and each described alignment settings comprises the setting of the second control loop fader and the second control loop phase adjuster settings.

4. feed forward amplifier system as claimed in claim 1 is characterized in that, the described parameter set that characterizes the condition of work of feed forward amplifier system comprises one or more temperature, DC power supply, input signal power and input signal carrier frequency.

5. feed forward amplifier system as claimed in claim 4 is characterized in that, the described parameter set that characterizes the condition of work of feed forward amplifier system is defined as attribute vector, and defines distance between any two attribute vectors.

6. feed forward amplifier system as claimed in claim 5, it is characterized in that, the described device that is used to control obtains current attribute vector and calculates the distance of the attribute vector of list element, and the list element that selection has minimum range is as the alignment settings in the controlled function.

7. feed forward amplifier system as claimed in claim 6 is characterized in that, the described device that is used to control is constantly measured the not calibration of feed forward amplifier system, when the not calibration of described measurement surpasses predetermined value, retrieves alignment settings from described alignment lists.

8. feed forward amplifier system as claimed in claim 7 is characterized in that, the described device that is used to control adopts selected element to set as initial calibration, and adopts iterative control algorithm to calculate new alignment settings from initial setting.

9. feed forward amplifier system as claimed in claim 8 is characterized in that, the described device that is used to control upgrades described alignment lists with new alignment settings after finishing described iterative computation.

10. feed forward amplifier system as claimed in claim 5 is characterized in that the distance between the immediate list element attribute vectors changes in whole tabulation.

11. an adaptive controller that is used for the loop of control amplifier system comprises:

One or more inputs are used to receive the one or more property parameters corresponding to the work at present condition of amplifier system; And

One or more processors, be coupled to described one or more input, and has relevant alignment lists, and with alignment list algorithm and controller algorithm programming, thereby provide the loop of loop adjustment setting with the control amplifier system, wherein said alignment list algorithm produces has the described tabulation that adjusting that described controller algorithm calculates is set, and one or more property parameters are associated with each adjusting setting.

12. the adaptive controller that is used for the loop of control amplifier system as claimed in claim 11, it is characterized in that, described alignment list algorithm is selected alignment settings from described alignment lists, use when starting or when loop becomes fully imbalance for described controller algorithm.

13. the adaptive controller that is used for the loop of control amplifier system as claimed in claim 12, it is characterized in that, described alignment list algorithm is by calculating corresponding to the distance between one or more property parameters of work at present condition and the property parameters relevant with each alignment settings in the tabulation, and select alignment settings corresponding to property parameters with minimum range, select alignment lists to regulate and set, use for described controller algorithm.

14. the adaptive controller that is used for the loop of control amplifier system as claimed in claim 12 is characterized in that distance calculation is come weighting to different property parameters with different weights.

15. the adaptive controller that is used for the loop of control amplifier system as claimed in claim 11 is characterized in that, described property parameters comprises temperature, DC supply voltage, input signal power and input signal carrier frequency wherein or multinomial.

16. the adaptive controller that is used for the loop of control amplifier system as claimed in claim 14 is characterized in that, between two sets of attribute parameters " n " and " 0 " apart from d _AttrL by weighting _InfThe L of norm distance metric or weighting ₂The norm distance metric defines.

17. the adaptive controller that is used for the loop of control amplifier system as claimed in claim 11 is characterized in that, also comprises the one or more inputs that are used to receive calibration data.

18. the adaptive controller that is used for the loop of control amplifier system as claimed in claim 17 is characterized in that, the described one or more inputs that are used to receive calibration data comprise the steering signal input.

19. the adaptive controller that is used for the loop of control amplifier system as claimed in claim 17 is characterized in that, the described one or more inputs that are used to receive calibration data comprise the input that is used for loop test data.

20. the adaptive controller that is used for the loop of control amplifier system as claimed in claim 11 is characterized in that, described adjusting is set and is comprised fader and phase adjuster settings.

21. method that is used for the control amplifier system, described amplifier system has the control loop that comprises control loop input, first signal path, secondary signal path and control loop output, in described first and second signal paths at least one comprises amplifier, and described method comprises:

List of alignment settings is provided, and each alignment settings has relevant condition of work;

The work at present condition of detecting amplifier system;

In work at present condition and the list of alignment settings those are compared; And

Select with tabulate in the most similar relevant alignment settings of condition of work.

22. the method that is used for the control amplifier system as claimed in claim 21 is characterized in that the related work condition is configured to the multidimensional property vector.

23. the method that is used for the control amplifier system as claimed in claim 22 is characterized in that, the described distance that relatively comprises between each attribute vector in current attribute vector of tolerance and the tabulation.

24. the method that is used for the control amplifier system as claimed in claim 23 is characterized in that, described selection comprises to be determined and the minimum attribute vector of the distance of work at present conditional attribute vector.

25. the method that is used for the control amplifier system as claimed in claim 21, it is characterized in that, also comprise and adopt iteration loop controller algorithm to calculate new alignment settings, wherein relevant with the most similar condition of work alignment settings is used as the initial calibration of adaptive loop circuit controller algorithm and sets.

26. the method that is used for the control amplifier system as claimed in claim 25 is characterized in that, comprises that also the new alignment settings that adopts the adaptive loop circuit controller algorithm to be calculated upgrades alignment lists.

27. the method that is used for the control amplifier system as claimed in claim 21 is characterized in that, the size of alignment lists is dynamic.

28. the method that is used for the control amplifier system as claimed in claim 23 is characterized in that, is changed in whole tabulation by the interval of institute's stored adjustment settings of attribute vector distance definition.

29. the method that is used for the control amplifier system as claimed in claim 28 is characterized in that, calibration is set the adjusting that higher density is set in the most responsive zone of the one or more conditions of work that comprise attribute vector in tabulation.

30. the method for the list of alignment settings of a control loop of safeguarding amplifier system, described tabulation comprises a plurality of elements, and each element has alignment settings and corresponding to the parameter set of the condition of work of amplifier system, described method comprises:

Select the element of alignment lists;

Determine to have the element of the alignment lists of the relevant work condition the most similar to selected element;

Determine whether that two elements are enough similar and be considered to redundant; And

If two elements of alignment lists are redundant, then delete element the earliest in these two elements.

31. the method for the list of alignment settings of the control loop of safeguarding amplifier system as claimed in claim 30 is characterized in that, the step of the element of described selection alignment lists comprises that before not being subjected in the selective listing tabulates and safeguards the element of handling the earliest.

32. the method for the list of alignment settings of the control loop of safeguarding amplifier system as claimed in claim 30, it is characterized in that the step that has the relevant work conditions of elements the most similar to selected element in described definite alignment lists comprises the distance metric of condition of work parameter value of each all the other element that is determined to alignment lists and the element that selection has minimum range.

33. the method for the list of alignment settings of the control loop of safeguarding amplifier system as claimed in claim 32 is characterized in that described distance metric comprises corresponding to the weighted difference between the parameter value of condition of work.

34. the method for the list of alignment settings of the control loop of safeguarding amplifier system as claimed in claim 33, it is characterized in that, comprise temperature, DC power supply, input signal power and input signal carrier frequency wherein or multinomial corresponding to the described parameter of the condition of work of amplifier system.

35. the method for the list of alignment settings of the control loop of safeguarding amplifier system as claimed in claim 30, it is characterized in that whether enough similar described definite element and be considered to redundant step and comprise the distance metric determined between the alignment settings and calibrated distance is compared with redundant distance threshold.

36. the method for the list of alignment settings of the control loop of safeguarding amplifier system as claimed in claim 32, it is characterized in that whether enough similar described definite element and be considered to redundant step and comprise the distance between the condition of work parameter of two elements is compared with expired distance threshold.

37. the method for the list of alignment settings of the control loop of safeguarding amplifier system as claimed in claim 30 is characterized in that, comprises that also repeating described tabulation for each element in the alignment lists safeguards and handle.

38. the method for the grading list of the alignment settings of a control loop that produces amplifier system, described tabulation comprises a plurality of elements, each element has alignment settings and corresponding to the relevant parameter collection of the condition of work of amplifier system, described tabulation has and comprises the hierarchy of two-stage at least, and described method comprises:

Element in the first order of selection alignment lists;

Determine to have the relevant work conditions of elements the most similar in the first order of alignment lists to selected element; And

The earliest element in two elements is downgraded to the more rudimentary of hierarchical alignment list.

39. the method for the grading list of the alignment settings of the control loop of generation amplifier system as claimed in claim 38, it is characterized in that, the step that has the relevant work conditions of elements the most similar to selected element in described definite alignment lists comprises the distance metric of condition of work of each all the other elements of the first order that is determined to alignment lists, and selects element with minimum range.

40. the method for the grading list of the alignment settings of the control loop of generation amplifier system as claimed in claim 38, it is characterized in that, also comprise determining whether two elements are redundant, wherein said element is early just just demoted under these elements are redundant situation.

41. the method for the grading list of the alignment settings of the control loop of generation amplifier system as claimed in claim 38 is characterized in that, comprises that also each level for grading list repeats described tabulation processing.

42. the method for the grading list of the alignment settings of the control loop of generation amplifier system as claimed in claim 41 is characterized in that, if the lowermost level of handling at hierarchy is safeguarded in tabulation, then deletes described clauses and subclauses early.

43. the method for the grading list of the alignment settings of the control loop of generation amplifier system as claimed in claim 38 is characterized in that, described degraded element is as the not next association of subset list entry of degraded redundant elements.

44. the method for the grading list of the alignment settings of the control loop of generation amplifier system as claimed in claim 43 is characterized in that, is demoted and subset list with the element of subset list and degraded redundant elements not merges.

45. method that is used for the control amplifier system, described amplifier system has the control loop that comprises control loop input, first signal path, secondary signal path and control loop output, in described first and second signal paths at least one comprises amplifier, and described method comprises:

Provide to have the grading list of the alignment settings of two-stage at least, each alignment settings have a alignment settings in the related work condition and the superlative degree part or all have subset alignment settings in more rudimentary;

The work at present condition of detecting amplifier system;

In the superlative degree of the grading list of work at present condition and alignment settings those are compared;

In the superlative degree of tabulation, select the alignment settings relevant with the most similar condition of work;

Those conditions of work in the subclass of comparison work at present condition and selected highest alignment settings;

Selection has the alignment settings of the subclass of similar condition of work; And

Be chosen in the alignment settings that has the condition of work the most similar in higher or more rudimentary to the work at present condition.

46. the method that is used for the control amplifier system as claimed in claim 45 is characterized in that, also comprises each grade reprocessing to grading list, is empty up to the next one than lower subsets.

47. the method that is used for the control amplifier system as claimed in claim 45 is characterized in that, the superlative degree has more rudimentary more rough alignment settings at interval.

48. the method that is used for the control amplifier system as claimed in claim 47 is characterized in that, any two alignment settings have calibrated distance, and highest have than the calibrated distance between the described more rudimentary bigger setting.

49. the method that is used for the control amplifier system as claimed in claim 47 is characterized in that, described calibrated distance is the weighted difference between the adjacent setting.

50. the method that is used for the control amplifier system as claimed in claim 49 is characterized in that, regulating setting is that gain-adjusted and phase adjusted are set, and weighting is the calibration susceptibility.

Images