CN102400902A - Method for evaluating reliability of performance state of reciprocating compressor - Google Patents

Abstract

The invention relates to a method for evaluating the reliability of the performance state of a reciprocating compressor. The method comprises the following steps of: determining a factor set U, wherein the factor set is a set formed by taking various factors influencing an evaluated object as elements, and the comprehensiveness, sensitivity, chanciness and stability of indexes are required to be taken into consideration when the indexes are selected; establishing a weight set A, and giving a corresponding weight to each factor, wherein the weight set is required to meet conditions of normalization and non-negativity; determining a remark set V, wherein the remark set is a classic set consisting of possible evaluation results of the evaluated object, the number m of remark grades is more than 3, but less than or equal to 9 and is an odd number; evaluating a single factor, and establishing a fuzzy relation matrix R; after the factor weight set A and the evaluation matrix R are determined, obtaining a fuzzy comprehensive evaluation set B according to the multiplying operation of the fuzzy matrix; and processing the evaluation indexes by a maximum membership grade method or a weighted averaging method. The method is simple and practical and is easy to popularize and apply.

Description

Reciprocal compressor performance state method for evaluating reliability

Technical field

The present invention relates to a kind of reciprocal compressor performance state method for evaluating reliability.

Background technique

Reciprocal compressor is because pressure range is wider; The thermal efficiency is higher, and adaptability is strong, and single stage compression is than high; In processing, conveying and other industrial department of a lot of oil, rock gas, occupy considerable status, its performance quality has very big influence to commercial production.In its using process, inevitably some faults can appear, influence running state, these have all brought serious hidden danger to safety in production.In case key equipment breaks down, not only enormous economic loss can be caused, and personal safety maybe be jeopardized, produce the important social influence.

Reciprocal compressor is done status monitoring and reliability evaluation, can avoid unnecessary disorderly closedown, reduce economic loss.How reducing the improper shutdown times of refinery process units is the matter of utmost importance that the refinery production management is paid close attention to.Existing practice shows; Status monitoring and reliability evaluation technology; Can be the ANOMALOUS VARIATIONS of discovering device state in the latent period of fault, help in time taking measures to prevent further expanding of equipment failure, reduce the improper shutdown times that the refinery device causes because of remarkable fault.According to statistics; Application state monitoring and reliability evaluation technology can not only minimizing accidents 75%, practice thrift maintenance man-hours 30%, practice thrift maintenance cost 25%-50%; Can also reduce production costs, energy saving and material consumption, greatly improve product quality and production efficiency.

The reliability evaluation of reciprocating compressor is the work of a more complicated.Factors such as people's experience, know-how and viewpoint; Often evaluation result there is bigger influence; And the multifactorial influence of the reliability audient of machinery; Continue to continue to use traditional reliability quantitative calculation method and come the reliability of calculating machine equipment relatively difficult nor enough scientific and reasonable, and adopt the reliability of more coming the assay machinery, can obtain comparatively ideal result near actual fuzzy theory.

Summary of the invention

The objective of the invention is to propose a kind of method of reciprocating compressor performance state reliability evaluation; Thermodynamic performance according to vibration parameters and unit operation; Set up the index system of reliability evaluation, adopt Field Using Fuzzy Comprehensive Assessment that compressor is carried out reliability evaluation, this method can be avoided excessively maintenance and correction maintenance to greatest extent; To the Field Force maintenance reference proposition is provided, guarantees refinery equipment safety, the operation of steady, good state.

The fuzzy evaluation principle that the present invention adopts is:

Choose evaluation index, confirm reciprocating compressor performance state evaluation indice.What at first consider is the good and bad information of compressor performance state that single index reflects, and fuzzy factors such as consideration assessor's experience and preference, representes the influence that evaluation index is good and bad to the compressor performance state with fuzzy evaluation matrix R.Different according to each evaluation index to the degree of compressor performance state reliability reflection, confirm its importance degree, be called weight.Fuzzy evaluation matrix R can calculate reciprocating compressor performance state reliability evaluation score and state grade with each index weight, and the Field Force can make corresponding operating in view of the above.

The method of a kind of reciprocating compressor performance state reliability evaluation of the present invention, concrete steps are gone into down: fuzzy overall evaluation comprises following six basic steps:

(1) confirms set of factors U

Set of factors is that the various factors of influence judge object is the set that element is formed, promptly

U＝(u ₁，u ₂，K，u _n)

Each element u _i(i=1,2, m) represent each influence factor, these factors can be blured, and also can be non-fuzzies.Confirm set of factors, also just constructed assessment indicator system.

To consider comprehensive, receptance, contingency, stability of index or the like during index for selection.See Fig. 1 reciprocating compressor group reliability evaluation index system arborescence.

(2) set up weight sets A

In order to reflect each factor u _i(i=1,2, L, significance level m) gives corresponding flexible strategy a to each factor _i(i=1,2, L, m), the collection that each flexible strategy is formed is called factor weight collection, i.e. weight sets.Be designated as

A＝(a ₁，a ₂，L，a _n)

Weight sets will satisfy normalization and nonnegativity condition:

1. gather among the U each index weight all greater than zero smaller or equal to 1, i.e. 0≤a _i≤1;

2. gather that each index weight algebraic sum equals 1 among the U,

See Fig. 2 reciprocating compressor group reliability evaluation index weight allocation figure.

(3) confirm comment collection V

The classics set that the comment collection is made up of the evaluation result that possibly make evaluation object can be expressed as

V＝(v ₁，v ₂，L，v _m)

In general, comment number of levels m is greater than 3 and be no more than 9, and m gets odd number more.This paper m=5, the comment collection is as shown in the table.

Table 2 comment rank

(4) carry out single factor evaluation, set up fuzzy relationship matrix r

$R=\left[\begin{array}{cccc}{r}_{11}& r_{12}& \mathrm{\Λ}& r_{1m}\\ r_{21}& r_{22}& \mathrm{\Λ}& r_{2m}\\ M& M& M& M\\ r_{n1}& n2& \mathrm{\Λ}& r_{\mathrm{nm}}\end{array}\right]$

Wherein, r _IjBe factor u among the U _iCorresponding V middle grade v _jMembership, just from factor u _iHave in mind and commented object can be cited as v _jThe membership of grade.r _IjBe i factor u _iTo the single factor evaluation of this things, it has constituted the basis of fuzzy overall evaluation.

This paper at first confirms the membership function of each index in reliability evaluation, with the corresponding membership function of each index value substitution to be evaluated, just can obtain evaluation vector then.At last altogether, just obtain the evaluation matrix (fuzzy relation matrix) of this sub-goal with each the index evaluation Vector Groups under the same sub-goal.

The membership function that adopts is:

${\mathrm{\&mu;}}_{\mathrm{ia}}\left(x\right)=\left\{\begin{array}{cc}1& (x\&le;a)\\ \frac{b-x}{b-a}& (a<x\&le;b)\\ 0& (x>b)\end{array}\right.$

${\mathrm{\&mu;}}_{\mathrm{ie}}\left(x\right)=\left\{\begin{array}{cc}0& (x\&le;d)\\ \frac{x-d}{e-d}& (d<x\&le;e)\\ 1& (x>e)\end{array}\right.$

μ _Iaμ _Ibμ _Icμ _Id, μ _IeRepresent that respectively index i is under the jurisdiction of the degree of a, b, c, d, each grade of e.

(5) fuzzy comprehensive evoluation

Factor weight collection A according to the multiplying of fuzzy matrix, obtains fuzzy comprehensive evoluation collection B after confirming with judge matrix R

$B=\mathrm{AoR}=(a_1,a_2,\mathrm{\&Lambda;},a_n)\left[\begin{array}{cccc}{r}_{11}& r_{12}& \mathrm{\&Lambda;}& r_{1m}\\ r_{21}& r_{22}& \mathrm{\&Lambda;}& r_{2m}\\ M& M& M& M\\ r_{n1}& n2& \mathrm{\&Lambda;}& r_{\mathrm{nm}}\end{array}\right]=(b_1,b_2,\mathrm{\&Lambda;},b_m)$

(5.1)

Wherein,

j=1; 2, L, m.

Judging quota b _jImplication be: taking all factors into consideration under the situation of all influence factors, passing judgment on the degree of membership of object for element j among the comment collection V.

(6) processing of judging quota

Obtaining judging quota b _j(j=1,2, L, m) after, need handle it, generally adopt maximum membership degree method or weighted mean method.What this paper adopted is weighted mean method.

The method for evaluating reliability that this paper proposes is compared with existing method and is had the following advantages:

1, quantifiable indicators such as the index involving vibrations parameter of the present invention's employing and thermal parameter have been abandoned the too strong and not strong index of receptance of ambiguity, are convenient to identification and judgement.

2, index system adopts modular construction, is convenient to management and upgrading; Both can carry out reliability evaluation, also can estimate each subtense angle and vitals to complete machine.

3, tentatively formulated the evaluation of indexes standard, advantage is the expert's marking that need not through loaded down with trivial details, can obtain each evaluation of indexes vector.

4, because the present invention uses fuzzy assessment method that the reciprocal compressor cylinder is carried out reliability evaluation, evaluation result for the scene provides effective reliable units operational data, has guaranteed the operation of safe, steady, the high-quality of refinery equipment accurately and reliably.

5, evaluating method of the present invention is equally applicable to the reciprocal compressor cylinder performance evaluation of other similar applications; Also can be generalized to reliability evaluation to other parts of reciprocating compressor and system.

6, method of the present invention is simple and practical, is easy to apply.

Description of drawings

Fig. 1 reciprocating compressor group reliability evaluation index system arborescence.

Fig. 2 reciprocating compressor group reliability evaluation index weight allocation figure.

Embodiment

Through Lanzhou Petrochemical Company 1,200,000 t/a diesel oil hydrogenation reciprocal compressor C-1101A are carried out the performance state reliability evaluation, the implementation methods and the effect of this method is described below.

On March 4th, 2009, we carried out data capture to hydrogenation unit C1101A.A compressor emergency shutdown maintenance when going to the scene again in second day.Recognize that from the maintenance man locating stud in the 2 cylinder piston rod holders watt comes off, fall into knock out drum, cause big noise from 2 cylinder outlet valves.The effect of locating stud is that radial location watt is carried out in holder, and is little to piston acting influence, but from security consideration, the maintenance man has still installed new locating stud.The data that we gathered with on March 4th, 2009 are fault data, and whether checking can judge the fault of unit parts through reliability evaluation.

(1) ask the free-ended evaluation matrix of motor:

The sub-goal motor free end u of reciprocating compressor group ₁₁Comprise vibration u ₁₁₁With temperature u ₁₁₂These two indexs, the value to be evaluated of these two indexs is respectively 0.298m/s ²With 39.8 ℃.The free-ended vibration of motor and temperature evaluation criterion such as table 5.4:

Vibration of table 5.4 motor free end and temperature evaluation criterion

Utilize this evaluation criterion can obtain the free-ended vibration of motor and the temperature index membership function following:

${\mathrm{\&mu;}}_{\mathrm{ia}}\left(x\right)=\left\{\begin{array}{cc}1& (x\&le;0.30)\\ \frac{0.39-x}{0.39-0.30}& (0.30<x\&le;0.39)\\ 0& (x>0.39)\end{array}\right.$

${\mathrm{\&mu;}}_{\mathrm{ie}}\left(x\right)=\left\{\begin{array}{cc}0& (x\&le;0.61)\\ \frac{x-0.61}{1.21-0.61}& (0.61<x\&le;1.21)\\ 1& (x>1.21)\end{array}\right.$

${\mathrm{\&mu;}}_{\mathrm{ja}}\left(x\right)=\left\{\begin{array}{cc}1& (x\&le;30)\\ \frac{40-x}{40-30}& (30<x\&le;40)\\ 0& (x>40)\end{array}\right.$

${\mathrm{\&mu;}}_{\mathrm{je}}\left(x\right)=\left\{\begin{array}{cc}0& (x\&le;75)\\ \frac{x-75}{80-75}& (75<x\&le;80)\\ 1& (x>80)\end{array}\right.$

With free-ended vibration of motor and temperature value difference substitution to be evaluated membership function separately, can obtain evaluation vector separately, then two evaluation vectors combinations are just obtained the free-ended evaluation matrix of motor:

$R_{11}=\left(\begin{array}{ccccc}1& 0& 0& 0& 0\\ 0.02& 0.98& 0& 0& 0\end{array}\right)$

(2) fuzzy comprehensive evoluation

Can check in A by Fig. 2 ₁₁=(0.80.2), so

$B_{11}=A_{11}o{R}_{11}=\left(\begin{array}{cc}0.8& 0.2\end{array}\right)\left[\begin{array}{ccccc}1& 0& 0& 0& 0\\ 0.02& 0.98& 0& 0& 0\end{array}\right]=\left(\begin{array}{ccccc}0.8& 0.2& 0& 0& 0\end{array}\right)$

So just obtained the free-ended reliability overall merit vector of motor.By that analogy, can obtain the reliability overall merit vector of other sub-goals at the same level by same algorithm.

B ₁₂＝(0.460.54000)

B ₁₃＝(0.440.56000)

B ₁₄＝(0.450.300.2500)

With motor and bent axle sub-goal u ₁Each sub-goal of next stage combines, and obtains motor and bent axle u ₁The evaluation matrix:

$R_1=\left\{\begin{array}{ccccc}0.8& 0.2& 0& 0& 0\\ 0.46& 0.54& 0& 0& 0\\ 0.44& 0.56& 0& 0& 0\\ 0.45& 0.30& 0.25& 0& 0\end{array}\right\}$

Motor and bent axle sub-goal u ₁Reliability overall merit vector:

$B_1=A_{1}o{R}_1=\left(\begin{array}{cccc}0.11& 0.22& 0.44& 0.23\end{array}\right)\left(\begin{array}{ccccc}0.8& 0.2& 0& 0& 0\\ 0.46& 0.54& 0& 0& 0\\ 0.44& 0.56& 0& 0& 0\\ 0.45& 0.30& 0.25& 0& 0\end{array}\right)$

$=\left(\begin{array}{ccccc}0.49& 0.45& 0.06& 0& 0\end{array}\right)$

According to such recursion mode, finally can obtain the overall merit vector of all sub-goals and general objective.Wherein the overall merit vector of general objective reciprocating compressor group is:

B ₀＝(0.330.390.130.140.01)

(6) scoring of evaluation vector and grading are handled

Evaluation vector with the weighted mean method operation of marking, is obtained quantitative evaluation.Weight vector can be taken as: S=(95,85,70,50,20).

According to final evaluation must score value, can be following 5 grades with each target reliabilities grade classification, see table 5.5.

Table 5.5 reciprocating compressor reliability step is divided

Be designated as example with motor free end specific item and carry out reliability scoring and grading:

F ₁₁＝B ₁₁oS ^T＝(0.80.2000)(9585705020) ^T＝93

Following table has been listed the result that the reciprocating compressor C-1101A data that on March 4th, 2009 was gathered are carried out reliability evaluation:

Table 5.62009 4, on March reciprocating compressor group C1101A reliability evaluation result

Can be found out that by last table the 2nd cylinder outlet valve reliability evaluation must be divided into 57.3, estimation scale is " poor ", meets on-the-spot actual conditions, and promptly 2 cylinder outlet valves part possibly break down.Through this instance analysis, explain that the reciprocating compressor group reliability evaluation model that this paper sets up is more reasonable, have using value preferably.

Claims (1)

1. reciprocal compressor performance state method for evaluating reliability is characterized in that:

(1) confirms set of factors U

Set of factors is that the various factors of influence judge object is the set that element is formed, promptly

U＝(u ₁，u ₂，K，u _n)

Each element u _i(i=1,2, L m) represents each influence factor, and these factors can be blured, and can be non-fuzzies also, have confirmed set of factors, have also just constructed assessment indicator system;

To consider comprehensive, receptance, contingency, the stability of index during index for selection;

(2) set up weight sets A

In order to reflect each factor u _i(i=1,2, L, significance level m) gives corresponding flexible strategy a to each factor _i(i=1,2, L, m), the collection that each flexible strategy is formed is called the factor weight collection, and promptly weight sets is designated as

A＝(a ₁，a ₂，L，a _n)

Weight sets will satisfy normalization and nonnegativity condition:

1. gather among the U each index weight all greater than zero smaller or equal to 1, i.e. 0≤a _i≤1;

2. gather that each index weight algebraic sum equals 1 among the U,

(3) confirm comment collection V

The classics set that the comment collection is made up of the evaluation result that possibly make evaluation object can be expressed as

V＝(v ₁，v ₂，L，v _m)

Comment number of levels m is greater than 3 and be no more than 9, and m gets odd number more;

(4) carry out single factor evaluation, set up fuzzy relationship matrix r

$R=\left[\begin{array}{cccc}{r}_{11}& r_{12}& \mathrm{\&Lambda;}& r_{1m}\\ r_{21}& r_{22}& \mathrm{\&Lambda;}& r_{2m}\\ M& M& M& M\\ r_{n1}& n2& \mathrm{\&Lambda;}& r_{\mathrm{nm}}\end{array}\right]$

Wherein, r _IjBe factor u among the U _iCorresponding V middle grade v _jMembership, just from factor u _iHave in mind and commented object can be cited as v _jThe membership of grade; r _IjBe i factor u _iTo the single factor evaluation of this things, it has constituted the basis of fuzzy overall evaluation;

At first confirm the membership function of each index in reliability evaluation, with the corresponding membership function of each index value substitution to be evaluated, just can obtain evaluation vector then; With each the index evaluation Vector Groups under the same sub-goal altogether, just obtain the evaluation matrix of this sub-goal, fuzzy relation matrix at last;

The membership function that adopts is:

${\mathrm{\&mu;}}_{\mathrm{ia}}\left(x\right)=\left\{\begin{array}{cc}1& (x\&le;a)\\ \frac{b-x}{b-a}& (a<x\&le;b)\\ 0& (x>b)\end{array}\right.$

${\mathrm{\&mu;}}_{\mathrm{ie}}\left(x\right)=\left\{\begin{array}{cc}0& (x\&le;d)\\ \frac{x-d}{e-d}& (d<x\&le;e)\\ 1& (x>e)\end{array}\right.$

μ _Ia, μ _Ib, μ _Ic, μ _Id, μ _IeRepresent that respectively index i is under the jurisdiction of the degree of a, b, c, d, each grade of e;

(5) fuzzy comprehensive evoluation

Factor weight collection A according to the multiplying of fuzzy matrix, obtains fuzzy comprehensive evoluation collection B after confirming with judge matrix R

$B=\mathrm{AoR}=(a_1,a_2,\mathrm{\&Lambda;},a_n)\left[\begin{array}{cccc}{r}_{11}& r_{12}& \mathrm{\&Lambda;}& r_{1m}\\ r_{21}& r_{22}& \mathrm{\&Lambda;}& r_{2m}\\ M& M& M& M\\ r_{n1}& n2& \mathrm{\&Lambda;}& r_{\mathrm{nm}}\end{array}\right]=(b_1,b_2,\mathrm{\&Lambda;},b_m)$

(5.1)

Wherein,

j=1; 2, L, m;

Judging quota b _jImplication be: taking all factors into consideration under the situation of all influence factors, passing judgment on the degree of membership of object for element j among the comment collection V;

(6) processing of judging quota

Obtaining judging quota, b _j(j=1,2, L, m) after, need handle it, adopt maximum membership degree method or weighted mean method.

Images