CN113300709A - Data processing algorithm - Google Patents

Abstract

The invention discloses a data processing algorithm, wherein arrays Q, N analog-to-digital conversion values of each array and corresponding power k of a calculated base number X are set; output array G₁、G₂、G₃、…G_QThe N analog-to-digital conversion values of each array are sequentially marked as G₁(AD_1，1、AD_1，2、…、AD_{1，（N‑1）}、AD_1，N)、G₂(AD_2，1、AD_2，2、…、AD_{2，（N‑1）}、AD _2，N)、…G_Q(AD_Q，1、AD_Q，2、…、AD_{Q，（N‑1）}、AD_Q，N) (ii) a S3, measuring a median value; a measure of the minimum; the maxima, modes, harmonic means, geometric means, and multi-dimensional metrics of each array are calculated. The data processing algorithm can remarkably improve the analog-to-digital conversion power by measuring the median, the minimum, the maximum, the mode, the harmonic mean and the arithmetic mean of a plurality of groups of analog-to-digital conversion values and then carrying out multi-dimensional measurement operation on the measurement valuesCompared with the existing ADC chip, the analog-to-digital conversion precision of the ADC chip is greatly improved.

Description

Data processing algorithm

Technical Field

The invention relates to a data processing algorithm of an analog-digital conversion circuit and a chip, in particular to a data processing algorithm.

Background

The current analog-to-digital conversion mode of the ADC chip is that an analog voltage value V and an analog-to-digital conversion output value are in a linear relationship:

analog-to-digital conversion value = (Grade ÷ V)_IC）× V_ADIf the decimal number is calculated, the integer can be obtained by 4-round-5-in method:

V_AD: simulating an input voltage value V;

V_IC: the reference voltage value, which is generally equal to the chip operating voltage, in most cases +5V, may also have a different value, but the chip is executing the programThe former must be determined in advance, i.e. a fixed value.

Grade: will V_ICEvenly divided into an equal number of parts, which may be 256, 4096, …, etc., but the chip must be determined in advance, i.e. a fixed value, before the program is executed.

At present, when ADC chip is used for actual analog-to-digital conversion, the same V is subjected to_ADA set of analog-to-digital converted values G (AD) obtained at different times₁、AD₂、…、AD_n-1、AD_n) Often inconsistent. The following is even often the case:

（1）V_AD1 < V_AD2but the two voltage values differ by a relatively small amount, and V_AD1Analog-to-digital conversion value of> V_AD2The analog-to-digital conversion value of (1);

（2）V_ADanalog-to-digital conversion value of voltage of = 0, 0>0 is a nonnegative integer value such as 1, 2, 3, 4, 5, etc.

Generally, it is common practice to use the array G average (from a mathematical point of view, it is also a linear process in practice) as the finally determined analog-to-digital conversion value, but the result of this method tends to reduce the accuracy of the chip analog-to-digital conversion.

Disclosure of Invention

The invention aims to provide a novel nonlinear measurement mode, compared with the linear relation of the traditional classical analog-to-digital conversion, the measurement accuracy is greatly improved, and the method is particularly helpful for detecting trace substances.

In order to achieve the above purpose, the following scheme is provided:

a data processing algorithm comprising the steps of:

s1, setting arrays Q, N analog-to-digital conversion values of each array and the corresponding power k of the calculation base number X;

s2 output array G₁、G₂、G₃、…、G_QEach array is N analog-to-digital conversion values, which are sequentially marked as G₁(AD_1,1、AD_1,2、…、AD_1,(N-1)、AD_1,N)、G₂(AD_2,1、AD_2,2、…、AD_2,(N-1)、AD _2,N)、G₃(AD_3,1、AD_3,2、…、AD_3,(N-1)、AD_3,N)、…、G_Q(AD_Q,1、AD_Q,2、…、AD_Q,(N-1)、AD_Q,N)；

S3, calculating G₁To G_QMedian measure values for each array;

s4, calculating G₁To G_QThe minimum value measurement value of each array;

s5, calculating G₁To G_QThe maximum value measurement value of each array;

s6, calculating G₁To G_QThe mode measurement value of each array;

s7, calculating G₁To G_QMeasuring the harmonic mean value of each array;

s8, calculating G₁To G_QMeasuring the geometric mean value of each array;

s9, calculating G₁To G_QMultidimensional measurements of each array.

Further, G in the step S3₁To G_QThe operation process of the median measurement value of each array comprises the following steps:

A1. get G₁To G_QMedian X of each array_1,i(i =1, 2, 3, ·, Q) are respectively denoted as X_1,1、X_1,2、 X_1,3、…X_1,Q；

G₁Median number X of_1,1Calculating, will G₁The output values of the analog-to-digital conversion are rearranged in order of magnitude: AD₁ ≤ AD₂ ≤…≤ AD_N-1 ≤AD_N,

If N is odd, then the median X_1,1 = AD_i，i=（N + 1）÷ 2,

If N is an even number, then the median X_1,1=（AD_i + AD_i+1）÷ 2，i=N÷2,

G₂To G_QMedian number X of_1,2To X_1,NValue process and G₁The consistency is achieved;

A2. calculation of G₁To G_QMedian measure for each array: p_M(M) = X_1,1 ^k+ X_1,2 ^k + … + X_1,Q ^k。

Further, G in the step S4₁To G_QThe operation process of the minimum value measurement value of each array comprises the following steps:

B1. get G₁To G_QMinimum value X of each array_2,i (i =1, 2, 3, ·, Q), each denoted X_2,1、X_2,2、 X_2,3、…X_2,Q:

X_2,i≤ AD_i,j (1 ≤ i ≤ Q ，j=1，2，3，···，N);

B2. Calculation of G₁To G_QMinimum value measurement values of each array: p_L(L) = X_2,1 ^k+ X_2,2 ^k + … + X_2,Q ^k。

Further, G in the step S5₁To G_QThe operation process of the maximum value measurement value of each array comprises the following steps:

C1. get G₁To G_QMaximum value X of each array_3,i(i =1, 2, 3, ·, Q), each denoted X_3,1、X_3,2、 X_3,3、…X_3,Q;

X_3,i≥ AD_i,j (1 ≤ i ≤ Q ，j=1，2，3，···，N),

C2. Calculation of G₁To G_QMaximum value measurement values of each array: p_H(H) = X_3,1 ^k+ X_3,2 ^k + … + X_3,Q ^k。

Further, G in the step S6₁To G_QThe operation process of the mode measurement value of each array comprises the following steps:

D1. get G₁To G_QMode X of each array_4,i(i =1, 2, 3, ·, Q), each denoted X_4,1、X_4,2、 X_4,3、…X_4,Q，

If there are r different values, in the same array G_j(j is not less than 1 and not more than Q), heavyThe number of repeated occurrences is the same as or the maximum, and then the mode X_4,jEqual to the arithmetic mean of r different values,

if in the same array G_j(j is more than or equal to 1 and less than or equal to Q), each number only appears once, then X_4,jEqual to the arithmetic mean of the array;

D2. calculation of G₁To G_QMeasurement of the mode of each array: p_S(S) = X_4,1 ^k+ X_4,2 ^k + … + X_4,Q ^k。

Further, G in the step S7₁To G_QThe operation process of the harmonic mean measuring value of each array comprises the following steps:

E1. get G₁To G_QHarmonic mean X for each array_5,i(i =1, 2, 3, ·, Q), each denoted X_5,1、X_5,2、 X_5,3、…X_5,Q，

If for G_iAny one of AD_i,jAre both greater than 0, (1. ltoreq. i.ltoreq. Q, j =1, 2, 3, N);

then X_5,i= N ÷ (1 ÷ AD ÷ or greater than N ÷_i,1+1÷AD_i,2+…+1÷AD_i,N)]The smallest integer of (a);

if for G_iAt least one AD being present_i,jEqual to 0 and one AD_i,jGreater than 0 (1. ltoreq. i.ltoreq. Q, j =1, 2, N), then X_5,i= 1；

If AD_i,jIs not equal to 0 (1 ≦ i ≦ Q, j =1, 2, 3 ·, N), then X_5,i = 0；

E2. Calculation of G₁To G_QHarmonic mean measure for each array: p_R(R) = X_5,1 ^k+ X_5,2 ^k + … + X_5,Q ^k。

Further, G in the step S8₁To G_QThe operation process of the geometric mean measurement value of each array comprises the following steps:

F1. get G₁To G_QMaximum value X of each array_6,i (i=1，2，3，···，Q), each occurrence is X_6,1、X_6,2、 X_6,3、…X_6,Q；

If for G_iAny one of AD_i,jAre all greater than 0, (1. ltoreq. i.ltoreq. Q, j =1, 2, 3, N),

then X_6,i= [ greater than or equal to (AD)_i,1•AD_i,2···AD_i,N)^1/N]The smallest integer of (a);

if for G_iAt least one AD being present_i,jEqual to 0 and one AD_i,jGreater than 0 (1. ltoreq. i.ltoreq. Q, j =1, 2, N), then X_6,i= 1；

If for G_i，AD_i,jIs ≡ 0 (1 ≦ i ≦ Q, j =1, 2, 3, N), then X_6,i= 0；

F2. Calculation of G₁To G_QGeometric mean measurement value P of each array_J(J) = X_6,1 ^k+ X_6,2 ^k + … + X_6,Q ^k。

Further, G in the step S9₁To G_QThe operation process of the multidimensional measurement value of each array comprises the following steps:

H1. select G₁To G_QThe following steps:

median X_1,iArray Q of₁（X_1,1、X_1,2、 X_1,3、…X_1,Q）；

Minimum value X_2,iArray Q of₂（X_2,1、X_2,2、 X_2,3、…X_2,Q）；

Maximum value X_3,iArray Q of₃（X_3,1、X_3,2、 X_3,3、…X_3,Q）；

X of the mode_4,iArray Q of₄（X_4,1、X_4,2、 X_4,3、…X_4,Q）；

X of harmonic mean_5,iArray Q of₅（X_5,1、X_5,2、 X_5,3、…X_5,Q）；

Geometric flatAverage number X_6,iArray Q of₆（X_6,1、X_6,2、 X_6,3、…X_6,Q）；

H2. Calculation of G₁To G_QMeasuring value of multidimensional measure:

P_T(T) = P_M(M)•P_L(L)•P_H(H)•P_S(S)•P_R(R)•P_J(J)

= (X_1,1 ^k+ X_1,2 ^k + … + X_1,Q ^k)•(X_2,1 ^k+ X_2,2 ^k + … + X_2,Q ^k)•(X_3,1 ^k+ X_3,2 ^k + … + X_3,Q ^k)

•(X_4,1 ^k+ X_4,2 ^k + … + X_4,Q ^k)•(X_5,1 ^k+ X_5,2 ^k + … + X_5,Q ^k)•(X_6,1 ^k+ X_6,2 ^k + … + X_6,Q ^k)。

the working principle and the advantages of the invention are as follows: the data processing algorithm measures the median, the minimum, the maximum, the mode, the harmonic mean and the arithmetic mean of a plurality of groups of analog-to-digital conversion values and then performs multidimensional measurement operation on the measures, so that the analog-to-digital conversion precision of an ADC chip of the analog-to-digital conversion circuit can be remarkably improved, and the operation precision is greatly improved compared with that of the conventional ADC chip.

Detailed Description

The following is further detailed by the specific embodiments:

the analog-digital conversion circuit is composed of two resistors R₁、R₀And ADC chip, R₁And R₀The two resistors are connected in series, the ADC chip is connected at the point of the series connection of the two resistors, the point of the series connection of the two resistors is marked as a point B, and the voltage V of the point B is_ADResistance R₁The other end of (1) is denoted as point A, and the resistance R₀And the other end of the same is grounded.

A data processing algorithm comprising the steps of:

s1, setting arrays Q, N analog-to-digital conversion values of each array and the corresponding power k of the calculation base number X;

s2 output array G₁、G₂、G₃、…、G_QEach array is N analog-to-digital conversion values, which are sequentially marked as G₁(AD_1,1、AD_1,2、…、AD_1,(N-1)、AD_1,N)、G₂(AD_2,1、AD_2,2、…、AD_2,(N-1)、AD _2,N)、G₃(AD_3,1、AD_3,2、…、AD_3,(N-1)、AD_3,N)、…、G_Q(AD_Q,1、AD_Q,2、…、AD_Q,(N-1)、AD_Q,N)；

S3, calculating G₁To G_QMedian measure values for each array;

s4, calculating G₁To G_QThe minimum value measurement value of each array;

s5, calculating G₁To G_QThe maximum value measurement value of each array;

s6, calculating G₁To G_QThe mode measurement value of each array;

s7, calculating G₁To G_QMeasuring the harmonic mean value of each array;

s8, calculating G₁To G_QMeasuring the geometric mean value of each array;

s9, calculating G₁To G_QMultidimensional measurements of each array.

G in the step S3₁To G_QThe operation process of the median measurement value of each array comprises the following steps:

A1. get G₁To G_QMedian X of each array_1,i(i =1, 2, 3, ·, Q) are respectively denoted as X_1,1、X_1,2、 X_1,3、…X_1,Q；

G₁Median number X of_1,1Calculating, will G₁The output values of the analog-to-digital conversion are rearranged in order of magnitude: AD₁ ≤ AD₂ ≤…≤ AD_N-1 ≤AD_N,

If N is odd, then the median X_1,1 = AD_i，i=（N + 1）÷ 2,

If N is an even number, then the median X_1,1=（AD_i + AD_i+1）÷ 2，i=N÷2,

G₂To G_QMedian number X of_1,2To X_1,NValue process and G₁The consistency is achieved;

A2. calculation of G₁To G_QMedian measure for each array: p_M(M) = X_1,1 ^k+ X_1,2 ^k + … + X_1,Q ^k。

Is formed by

The sigma-domain is generated in such a way that,

is a measurable space, P_M(M) is a median measure of the measurable space. According to the theory of mathematical metric theory, the specific structure of the measurement space and measurement is as follows:

(1)M＝{X_1,i：X_1,iis G_1,i(AD₁、AD₂、…、AD_N-1、AD_N) Q and N are any non-negative integer }, the median being defined:

will be array G₁、G₂、G₃…G_QThe analog-to-digital conversion numbers of the arrays are rearranged according to the size sequence: AD₁≤AD₂≤…≤AD_N-1≤AD_NIf N is an odd number, then x is AD_i，i＝(N+1)÷2，

If N is an even number, x is (AD)_i+AD_i+1)÷2，i＝N÷2，

If N is equal to 0, then M is equal to phi and is empty set,

for any two arrays G_iAnd G_jThe median of (i ≠ j) is X_iAnd X_jAll belong to M, if X_iIs equal to X_jStill considering the two median values as being different elements of the set M, namely X_i≠X_jThat is to say: different elements (median) in M may correspond to the same non-negative integer;

(2)Ω_M＝{X_1,i: all median }, any set M being a subset of Ω, i.e. M is a subset of Ω

(3) According to the theory of metrics, the median set is of the kind

I.e. to any

In that

Produce only one minimum sigma domain, noted

) Is a measurable space in which the measurement of the object,

in a measurable space (omega)_Mσ (M)) is defined as:

P_M(φ)＝0,

P_M(M)＝X_1,1 ^K+X_1,2 ^K+X_1,3 ^K+…+X_1,Q ^Kwherein

According to the theory of meteorology, P_M(M) is a measurable space

A measure of (d);

(4) according to the theory of measures P_M(M) A non-linear measure of the median set of the groups of chip analog-to-digital conversion output values corresponding to the voltage V at mid B_ADA non-linear relationship. The traditional classical analog-to-digital conversion value and voltage V_ADIs a linear relationship. The measurement can accurately distinguish small voltage differences;

(5) q, N, k can be determined according to the actual conditions of chip program memory and data memory size, operation speed, required accuracy, required completion time, etc., such as Q-128, N-32, k-2.

And median measure P_M(M) obtaining a minimum measurement value P_L(L), maximum value measurement value P_H(H) Measurement of the mode P_S(S) harmonic mean value P_R(R), geometric mean value of measurement P_J（J）。

G in the step S4₁To G_QThe operation process of the minimum value measurement value of each array comprises the following steps:

B1. get G₁To G_QMinimum value X of each array_2,i (i =1, 2, 3, ·, Q), each denoted X_2,1、X_2,2、 X_2,3、…X_2,Q:

X_2,i≤ AD_i,j (1 ≤ i ≤ Q ，j=1，2，3，···，N);

B2. Calculation of G₁To G_QMinimum value measurement values of each array: p_L(L) = X_2,1 ^k+ X_2,2 ^k + … + X_2,Q ^k。

Set of infinitesimal values L = { X =_2,i：X_2,iIs an array G_1，Q(AD₁、AD₂、…、AD_N-1、AD_N) Is a minimum value of (a), Q and N are any non-negative integer }

If N = 0, L = phi is an empty set,

Ω_L = { X_2,i: all minima, any set L is Ω_LI.e. L ⊂ Ω_L，

Definition of minimum value: x is less than or equal to AD_i，i = 1、2、…、N，

For any two arrays G_iAnd G_jThe minimum values of (i ≠ j) are X_iAnd X_jAll belong to L, if X_iIs equal to X_jThe two minima are still considered to be distinct elements in the set L, namely X_i≠X_jThat is, different elements (minima) in L may correspond to the same non-negative integer.

Minimum value set class

= L: l is a set of minimum values }, i.e., for any L ∈

。

G in the step S5₁To G_QThe operation process of the maximum value measurement value of each array comprises the following steps:

C1. get G₁To G_QMaximum value X of each array_3,1(i =1, 2, 3, ·, Q), each denoted X_3,1、X_3,2、 X_3,3、…X_3,Q;

X_3,i≥ AD_i,j (1 ≤ i ≤ Q ，j=1，2，3，···，N),

C2. Calculation of G₁To G_QMaximum value measurement values of each array: p_H(H) = X_3,1 ^k+ X_3,2 ^k + … + X_3,Q ^k。

Maximum value set H = { X_3,i：X_3,iIs an array G_1，Q (AD₁、AD₂、…、AD_N-1、AD_N) Is the maximum value of (a), Q and N are any non-negative integer }

Definition of maximum value: x is not less than AD_i,i = 1、2、…、N,

If N = 0, H = Φ is an empty set.

Ω_H = { X_3,i: all maxima }, any set H is Ω_HI.e. H ⊂ Ω_H，

For any two arrays G_iAnd G_j(i ≠ j) has a maximum value X_iAnd X_jAll belong to H, if X_iIs equal to X_jThe two maxima are still considered to be distinct elements in the set H, namely X_i≠X_jThat is to say: different elements (maxima) in H may correspond to the same non-negative integer.

Maximum set class:

= H: h is a set of maxima }, i.e., for any H ∈

。

G in the step S6₁To G_QThe operation process of the mode measurement value of each array comprises the following steps:

D1. get G₁To G_QMode X of each array_4,i(i =1, 2, 3, ·, Q), each denoted X_4,1、X_4,2、 X_4,3、…X_4,Q，

If there are r different values, in the same array G_j(j is more than or equal to 1 and less than or equal to Q), the repeated occurrence times are as many as the maximum repeated occurrence times, and the mode X is_4,jEqual to the arithmetic mean of r different values,

if in the same array G_j(j is more than or equal to 1 and less than or equal to Q), each number only appears once, then X_4,jEqual to the arithmetic mean of the array;

D2. calculation of G₁To G_QMeasurement of the mode of each array: p_S(S) = X_4,1 ^k+ X_4,2 ^k + … + X_4,Q ^k。

Mode set: s = { X_4,i：X_4,iIs an array G_1,Q (AD₁、AD₂、…、AD_N-1、AD_N) Q and N are any non-negative integer }

Definition of mode: x_4,iIs a corresponding array G_1,Q (AD₁、AD₂、…、AD_N-1、AD_N) The number of occurrences of the compound is the greatest;

if N = 0, S = Φ is an empty set.

Ω_S = {X_4,i: all modes }, any set S is a subset of Ω S, i.e., S ⊂ Ω_S

For any two arrays G_iAnd G_jThe mode of (i ≠ j) is X_iAnd X_jAll belong to S, if X_iIs equal to X_jStill considering the two modes as being distinct elements in set S, namely X_i≠X_jThat is to say: different elements (modes) in S may correspond to the same non-negative integer.

If there are r different values, in an array G_1,QThe number of repeated occurrences is as large as, and is the largest. X is equal to the arithmetic mean of the r different values.

If in array G_1,QIn, each number appears only once, then X_4,iIs equal to the array G_1,QThe arithmetic mean of (a) is calculated,

a collection of masses

= S: s is a set of modes }, i.e. for any set of S e

。

G in the step S7₁To G_QThe operation process of the harmonic mean measuring value of each array comprises the following steps:

E1. get G₁To G_QHarmonic mean X for each array_5,i(i =1, 2, 3, ·, Q), each denoted X_5,1、X_5,2、 X_5,3、…X_5,Q，

If for G_iAny one of AD_i,jAre both greater than 0, (1. ltoreq. i.ltoreq. Q, j =1, 2, 3, N);

then X_5,i= N ÷ (1 ÷ AD ÷ or greater than N ÷_i,1+1÷AD_i,2+…+1÷AD_i,N)]The smallest integer of (a);

if for G_iAt least one AD being present_i,jEqual to 0 and one AD_i,jGreater than 0 (1. ltoreq. i.ltoreq. Q, j =1, 2, N), then X_5,i= 1；

If AD_i,jIs not equal to 0 (1 ≦ i ≦ Q, j =1, 2, 3 ·, N), then X_5,i = 0；

E2. Calculation of G₁To G_QHarmonic mean measure for each array: p_R(R) = X_5,1 ^k+ X_5,2 ^k + … + X_5,Q ^k。

Harmonic mean set R = { X_5,i：X_5,iIs an array G_1,Q (AD₁、AD₂、…、AD_N-1、AD_N) Is equal to the harmonic mean of (1), Q and N are any non-negative integer }

Definition of harmonic mean:

if any AD_iAre both greater than 0, then X_5,i= greater than or equal to N ÷ ((1/AD)₁）+（1/AD₂）+（1/AD₃）+…+（1/AD_N) Minimum integer of (c)

If there is at least one AD_iEqual to 0 and one AD_jGreater than 0 (i is greater than or equal to 1, j is less than or equal to n) then X_5,i = 1，

If AD_iIs not less than 0 (1. ltoreq. i. ltoreq.n) then X_5,i = 0，

If N = 0, R = Φ is the empty set.

Ω_R = {X_5,i: all harmonic means }, any set R is Ω_RI.e. R ⊂ Ω_R

For any two arrays G_iAnd G_jThe harmonic mean of (i ≠ j) is X_iAnd X_jAll belong to R, if X_iIs equal to X_jThe two harmonic means are still considered to be different elements in the set R, namely X_i≠X_jThat is to say: multiple different elements in R (harmonic mean)Mean) may correspond to the same non-negative integer.

Harmonic mean collection of numbers

= R: r is a set of harmonic means }, i.e. for any R ∈

。

G in the step S8₁To G_QThe operation process of the geometric mean measurement value of each array comprises the following steps:

F1. get G₁To G_QMaximum value X of each array_6,1(i =1, 2, 3, ·, Q), each denoted X_6,1、X_6,2、 X_6,3、…X_6,Q；

If for G_iAny one of AD_i,jAre all greater than 0, (1. ltoreq. i.ltoreq. Q, j =1, 2, 3, N),

then X_6,i= [ greater than or equal to (AD)_i,1•AD_i,2···AD_i,N)^1/N]The smallest integer of (a);

if for G_iAt least one AD being present_i,jEqual to 0 and one AD_i,jGreater than 0 (1. ltoreq. i.ltoreq. Q, j =1, 2, N), then X_6,i= 1；

If for G_i，AD_i,jIs ≡ 0 (1 ≦ i ≦ Q, j =1, 2, 3, N), then X_6,i= 0；

F2. Calculation of G₁To G_QGeometric mean measurement value P of each array_J(J) = X_6,1 ^k+ X_6,2 ^k + … + X_6,Q ^k。

Collection of geometric mean: j = { X_6,i：X_6,iIs an array G_1,Q (AD₁、AD₂、…、AD_N-1、AD_N) A harmonic mean of), Q and N are any non-negative integers },

definition of harmonic mean:

if any AD_i Are all greater than 0 and are all greater than 0,then X_6,i= is greater than or equal to (X)_6,1·X_6,2·X_6,3·…·X_6,N）^1/NIs the smallest integer of (a) or (b),

if there is at least one AD_iEqual to 0 and one AD_jGreater than 0 (i is greater than or equal to 1, j is less than or equal to n) then X_6,1 = 1,

If AD_iIs not less than 0 (1. ltoreq. i. ltoreq.n) then X_6,1= 0,

If N = 0, J = Φ is an empty set.

Ω_J = { X_6,1: all geometric means }, any set J is Ω_JI.e. J ⊂ Ω_J,

For any two arrays G_iAnd G_jThe geometric mean of (i ≠ j) is X_iAnd X_jAll belong to J, if X_iIs equal to X_jThe two geometric means are still considered to be different elements in the set J, namely X_i≠X_jThat is to say: different elements (geometric means) in J may correspond to the same non-negative integer.

Class of geometric mean number set

= J: j is a set of geometric means }, i.e., for any J ∈

。

G in the step S9₁To G_QThe operation process of the multidimensional measurement value of each array comprises the following steps:

H1. select G₁To G_QThe following steps:

median X_1,iArray Q of₁（X_1,1、X_1,2、 X_1,3、…X_1,Q）；

Minimum value X_2,iArray Q of₂（X_2,1、X_2,2、 X_2,3、…X_2,Q）；

Maximum value X_3,iArray Q of₃（X_3,1、X_3,2、 X_3,3、…X_3,Q）；

X of the mode_4,iArray Q of₄（X_4,1、X_4,2、 X_4,3、…X_4,Q）；

X of harmonic mean_5,iArray Q of₅（X_5,1、X_5,2、 X_5,3、…X_5,Q）；

X of geometric mean_6,iArray Q of₆（X_6,1、X_6,2、 X_6,3、…X_6,Q）；

H2. Calculation of G₁To G_QMeasuring value of multidimensional measure:

P_T(T) = P_M(M)•P_L(L)•P_H(H)•P_S(S)•P_R(R)•P_J(J)

= (X_1,1 ^k+ X_1,2 ^k + … + X_1,Q ^k)•(X_2,1 ^k+ X_2,2 ^k + … + X_2,Q ^k)•(X_3,1 ^k+ X_3,2 ^k + … + X_3,Q ^k)

•(X_4,1 ^k+ X_4,2 ^k + … + X_4,Q ^k)•(X_5,1 ^k+ X_5,2 ^k + … + X_5,Q ^k)•(X_6,1 ^k+ X_6,2 ^k + … + X_6,Q ^k)。

according to the theory of mathematical metric theory, in each set class

、

、

、

、

、

、

The above can generate the smallest sigma domain, which is sequentially marked as sigma: (

)、σ(

)、σ(

)、σ(

)、σ(

)、σ(

)、σ(

) Thereby forming a measurable space (omega) from the median_M，σ(

) Minimum measurable space (omega)_L，σ(

) Maximum measurable space (omega)_H，σ(

) Etc. and mode measurable space (omega)_S，σ(

) Harmonic mean measurable space (Ω)_R，σ(

) Geometric mean measurable space (omega)_J，σ(

) A product measure space (Ω) can be constructed_T，σ(

) M, L, H, S, R, J, T are each a measurable set of these measurable spaces. The concrete structure is as follows:

(1) product set T = M × L × H × S × R × J { (X)₁, X₂, X₃, X₄, X₅, X₆)：X₁∈M, X₂∈L, X₃∈H, X₄∈S, X₅∈R, X₆∈J}，

(2) Product set omega_T=Ω_M×Ω_L×Ω_H×Ω_S×Ω_R×Ω_J，

(3) Collections class

=

×

×

×

×

×

={ M×L×H×S×R×J：M∈

, L∈

, H∈

, S∈

, R∈

, J∈

}，

As mentioned above, it is also possible to arbitrarily select sets and clusters that are products of sets and clusters,

(4) according to the theory of the metric theory, only one minimum sigma domain is generated at T and is marked as sigma: (

)，（Ω_T，σ(

)，P_T) Is (omega)_M，σ(

)，P_M）、（Ω_L，σ(

)，P_L）、（Ω_H，σ(

)，P_H）、（Ω_S，σ(

)，P_S）、（Ω_R，σ(

)，P_R）、（Ω_J，σ(

)，P_J) The independent product measure space of (a), the measure of which is:

P_T(T) = P(M×L×H×S×R×J)= P_M(M)•P_L(L)•P_H(H)•P_S(S)•P_R(R)•P_J(J)，

(5) according to the theory of measures P_T(T) can be used as a non-linear measure of the analog-to-digital conversion value of the chip, namely, the voltage V at the point B_ADThe method is different from the traditional classical analog-to-digital linear conversion, so that the small change of the voltage can be accurately distinguished.

(6) The unified Q, N, k values are set for the various different measurement calculations described above, and different Q values can be determined based on chip program memory, data memory, computation speed, required accuracy, completion time, etc_j、N_j、k_j(j =1 ~ 6).

The data processing algorithm measures the median, the minimum, the maximum, the mode, the harmonic mean and the arithmetic mean of a plurality of groups of analog-to-digital conversion values and then performs multidimensional measurement operation on the measures, so that the analog-to-digital conversion precision of an ADC chip of the analog-to-digital conversion circuit can be remarkably improved, and the operation precision is greatly improved compared with that of the conventional ADC chip.

The measurement is obviously different from the traditional classical linear analog-to-digital conversion mode related to voltage, but is a novel nonlinear measurement mode established according to the theory of mathematical measure theory, so that the small difference of voltage can be accurately distinguished, and the effect is greatly improved compared with the original precision.

The value of Q, N, K is determined by selecting one or more of the above measures according to chip program and data memory size, speed of operation, required accuracy and completion time, (e.g., Q = 128: representing 128 arrays, N = 32: representing chips each having 32 analog-to-digital converted values, and K = 2: representing the square of the corresponding value of the set element)

The rule of selection is as follows:

(1) the larger Q and N are, the larger the chip data memory is needed to be, otherwise, the smaller the chip data memory is needed to be;

(2) the larger K is, the longer the data processing time of the chip is, otherwise, the shorter the data processing time of the chip is;

(3) the more the selection measure types are, the longer the data processing time of the chip is, otherwise, the shorter the data processing time is;

(4) in short, the more the selection measure types are, the larger the value of Q, N, K is, the larger the chip program memory and the data memory are needed, the longer the data processing time is, and the higher the accuracy is; if the selection metric is smaller in kind and the value of Q, N, K is smaller, the chip program memory and the data memory are required to be smaller, the data processing time is shorter, and the accuracy is lowered.

The foregoing is merely an example of the present invention, and common general knowledge in the field of known specific structures and characteristics of the embodiments is not described herein in any greater extent than that known to persons of ordinary skill in the art at the filing date or before the priority date of the present invention, so that all of the prior art in this field can be known and can be applied with the ability of conventional experimental means before this date. It should be noted that, for those skilled in the art, without departing from the structure of the present invention, several changes and modifications can be made, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the applicability of the patent. The scope of the claims of the present application shall be determined by the contents of the claims, and the description of the embodiments and the like in the specification shall be used to explain the contents of the claims.

Claims (8)

1. A data processing algorithm, comprising the steps of:

s1, setting arrays Q, N analog-to-digital conversion values of each array and the corresponding power k of the calculation base number X;

s2 output array G₁、G₂、G₃、…、G_QEach array is N analog-to-digital conversion values, which are sequentially marked as G₁(AD_1,1、AD_1,2、…、AD_1,(N-1)、AD_1,N)、G₂(AD_2,1、AD_2,2、…、AD_2,(N-1)、AD _2,N)、G₃(AD_3,1、AD_3,2、…、AD_3,(N-1)、AD_3,N)、…、G_Q(AD_Q,1、AD_Q,2、…、AD_Q,(N-1)、AD_Q,N)；

S3, calculating G₁To G_QMedian measure values for each array;

s4, calculating G₁To G_QThe minimum value measurement value of each array;

s5, calculating G₁To G_QThe maximum value measurement value of each array;

s6, calculating G₁To G_QThe mode measurement value of each array;

s7, calculating G₁To G_QMeasuring the harmonic mean value of each array;

s8, calculating G₁To G_QMeasuring the geometric mean value of each array;

s9, calculating G₁To G_QMultidimensional measurements of each array.

2. The data processing algorithm of claim 1, wherein G in step S3₁To G_QThe operation process of the median measurement value of each array comprises the following steps:

A1. get G₁To G_QMedian X of each array_1,i(i =1, 2, 3, ·, Q) are respectively denoted as X_1,1、X_1,2、 X_1,3、…X_1,Q；

G₁Median number X of_1,1Calculating, will G₁The output values of the analog-to-digital conversion are rearranged in order of magnitude: AD₁ ≤ AD₂ ≤ … ≤ AD_N-1 ≤AD_N,

If N is odd, then the median X_1,1 = AD_i，i=（N + 1）÷ 2,

If N is an even number, then the median X_1,1=（AD_i + AD_i+1）÷ 2，i=N÷2,

G₂To G_QMedian number X of_1,2To X_1,NValue process and G₁The consistency is achieved;

A2. calculation of G₁To G_QMedian measure for each array: p_M(M) = X_1,1 ^k+ X_1,2 ^k + … + X_1,Q ^k。

3. The data processing algorithm of claim 1, wherein G in step S4₁To G_QThe operation process of the minimum value measurement value of each array comprises the following steps:

B1. get G₁To G_QMinimum value X of each array_2,i (i =1, 2, 3, ·, Q), each denoted X_2,1、X_2,2、 X_2,3、…X_2,Q:

X_2,i ≤ AD_i,j (1 ≤ i ≤ Q ，j=1，2，3，···，N);

B2. Calculation of G₁To G_QMinimum value measurement values of each array: p_L(L) = X_2,1 ^k+ X_2,2 ^k + … + X_2,Q ^k。

4. The data processing algorithm of claim 1, characterized in thatIn step S5, G₁To G_QThe operation process of the maximum value measurement value of each array comprises the following steps:

C1. get G₁To G_QMaximum value X of each array_3,i(i =1, 2, 3, ·, Q), each denoted X_3,1、X_3,2、 X_3,3、…X_3,Q;

X_3,i≥ AD_i,j (1 ≤ i ≤ Q ，j=1，2，3，···，N),

C2. Calculation of G₁To G_QMaximum value measurement values of each array: p_H(H) = X_3,1 ^k+ X_3,2 ^k + … + X_3,Q ^k。

5. The data processing algorithm of claim 1, wherein G in step S6₁To G_QThe operation process of the mode measurement value of each array comprises the following steps:

D1. get G₁To G_QMode X of each array_4,i(i =1, 2, 3, ·, Q), each denoted X_4,1、X_4,2、 X_4,3、…X_4,Q，

If there are r different values, in the same array G_j(j is more than or equal to 1 and less than or equal to Q), the repeated occurrence times are as many as the maximum repeated occurrence times, and the mode X is_4,jEqual to the arithmetic mean of r different values,

if in the same array G_j(j is more than or equal to 1 and less than or equal to Q), each number only appears once, then X_4,jEqual to the arithmetic mean of the array;

D2. calculation of G₁To G_QMeasurement of the mode of each array: p_S(S) = X_4,1 ^k+ X_4,2 ^k + … + X_4,Q ^k。

6. The data processing algorithm of claim 1, wherein G in step S7₁To G_QThe operation process of the harmonic mean measurement value of each array comprises the following stepsThe method comprises the following steps:

E1. get G₁To G_QHarmonic mean X for each array_5,i(i =1, 2, 3, ·, Q), each denoted X_5,1、X_5,2、 X_5,3、…X_5,Q，

If for G_iAny one of AD_i,jAre both greater than 0, (1. ltoreq. i.ltoreq. Q, j =1, 2, 3, N);

then X_5,i= N ÷ (1 ÷ AD ÷ or greater than N ÷_i,1+1÷AD_i,2+…+1÷AD_i,N)]The smallest integer of (a);

if for G_iAt least one AD being present_i,jEqual to 0 and one AD_i,jGreater than 0 (1. ltoreq. i.ltoreq. Q, j =1, 2, N), then X_5,i= 1；

If AD_i,jIs not equal to 0 (1 ≦ i ≦ Q, j =1, 2, 3 ·, N), then X_5,i = 0；

E2. Calculation of G₁To G_QHarmonic mean measure for each array: p_R(R) = X_5,1 ^k+ X_5,2 ^k + … + X_5,Q ^k。

7. The data processing algorithm of claim 1, wherein G in step S8₁To G_QThe operation process of the geometric mean measurement value of each array comprises the following steps:

F1. get G₁To G_QMaximum value X of each array_6,i(i =1, 2, 3, ·, Q), each denoted X_6,1、X_6,2、 X_6,3、…X_6,Q；

If for G_iAny one of AD_i,jAre all greater than 0, (1. ltoreq. i.ltoreq. Q, j =1, 2, 3, N),

then X_6,i= [ greater than or equal to (AD)_i,1•AD_i,2···AD_i,N)^1/N]The smallest integer of (a);

if for G_iAt least one AD being present_i,jEqual to 0 and one AD_i,jGreater than 0 (1 ≤ andi ≦ Q, j =1, 2, ·, N), then X_6,i= 1；

If for G_i，AD_i,jIs ≡ 0 (1 ≦ i ≦ Q, j =1, 2, 3, N), then X_6,i= 0；

F2. Calculation of G₁To G_QGeometric mean measurement value P of each array_J(J) = X_6,1 ^k+ X_6,2 ^k + … + X_6,Q ^k。

8. The data processing algorithm according to claims 1-7, wherein G in step S9₁To G_QThe operation process of the multidimensional measurement value of each array comprises the following steps:

H1. select G₁To G_QThe following steps:

median X_1,iArray Q of₁（X_1,1、X_1,2、 X_1,3、…X_1,Q）；

Minimum value X_2,iArray Q of₂（X_2,1、X_2,2、 X_2,3、…X_2,Q）；

Maximum value X_3,iArray Q of₃（X_3,1、X_3,2、 X_3,3、…X_3,Q）；

X of the mode_4,iArray Q of₄（X_4,1、X_4,2、 X_4,3、…X_4,Q）；

X of harmonic mean_5,iArray Q of₅（X_5,1、X_5,2、 X_5,3、…X_5,Q）；

X of geometric mean_6,iArray Q of₆（X_6,1、X_6,2、 X_6,3、…X_6,Q）；

H2. Calculation of G₁To G_QMeasuring value of multidimensional measure:

P_T(T) = P_M(M)•P_L(L)•P_H(H)•P_S(S)•P_R(R)•P_J(J)

= (X_1,1 ^k+ X_1,2 ^k + … + X_1,Q ^k)•(X_2,1 ^k+ X_2,2 ^k + … + X_2,Q ^k)•(X_3,1 ^k+ X_3,2 ^k + … + X_3,Q ^k)

•(X_4,1 ^k+ X_4,2 ^k + … + X_4,Q ^k)•(X_5,1 ^k+ X_5,2 ^k + … + X_5,Q ^k)•(X_6,1 ^k+ X_6,2 ^k + … + X_6,Q ^k)。