CN105928694A - Moving sieve fault diagnosis method based on Weibull analysis model - Google Patents

Abstract

The invention discloses a moving sieve fault diagnosis method based on a Weibull analysis model. The method is characterized by collecting impact force of coal falling off from a moving sieve plate on a buffer board through an impact force sensor, and calculating proportion range of the coal falling off from the moving sieve plate in the total amount of a sample when the moving sieve plate is normal, that is, coal size distribution range; and analyzing the coal falling process in the time period t, and according to the total impact force P collected by the impact force sensor, calculating the proportion of the total quantity of the coal falling off from the moving sieve plate within the time period t in the total quantity of the sample, and making a comparison between the proportion and the proportion range obtained when the moving sieve plate is normal to measure the total quantity of the falling-off coal and realize moving sieve fault diagnosis. The method can realize moving sieve fault diagnosis only through the impact force of the coal on the buffer board, can roughly measure damage degree of the moving sieve plate, is simple to realize and high in precision, does not depend on application scenarios and improves economic benefit and intelligent supervision level of a coal preparation plant.

Description

A kind of dynamic sieve method for diagnosing faults analyzing model based on Weibull

Technical field

The present invention relates to dynamic sieve method for diagnosing faults field, a kind of based on Weibull analysis model dynamic Sieve method for diagnosing faults.

Background technology

Dong Shaishi coal preparation plant separates the visual plant of coal cinder granularity.Owing to coal separation is a continuous print line production Process, equipment room coupling is relatively big, and equipment has strict requirements to pan feeding parameter, once goes wrong, By causing, machine follow-up in production procedure is impaired, results even in stopping production, therefore dynamic sieve fault prison time serious Survey problem becomes one of important topic, dynamic sieve carries out fault diagnosis fast, accurately and is related to the effect of coal separation Rate and safety, and simplify, promptness, degree of accuracy are to sieve the important indicator that diagnostic method is good and bad.

Fault diagnosis be monitoring object when breaking down can according to the state of current device and mechanism, around other The phenomenon of relatedness things, the feature etc. of fault are analyzed location of fault, degree and reason, and are sent police Accuse or other promptings, and then directive function is played in whole production work.Fault diagnosis seems outstanding within coal mines For important, not only it be related to the operation in whole colliery, also miner's lives and properties produced tremendous influence. Monitoring method for dynamic sieve mainly has at present: manual inspection, video surveillance, equipment feature analysis, but Video surveillance, equipment feature analysis due to environmental factors (coal dust, dust), working condition (noise) etc., Being substantially at conceptual phase at present, the widely used human at periodic intervals of being checks.

(1) manual inspection

Sieve plate, according to the experience cycle of damages of screen deck, is checked or changes by coal preparation plant workman in advance, the party Though method The simplest, but waste time and energy, it is impossible to accomplish that fault is monitored in real time.Patrol additionally, this method height relies on The experience of inspection workman, the foundation changed as part of appliance using it is the most unreliable, may waste sieve in a large number Plate.

(2) video surveillance

The method of video surveillance is at the sieve plate dust-proof video camera of high definition installed above, by monitor video image Processing, whether slot size damages, whether sieve plate ruptures in diagnosis.But practical situation sieve plate, coal cinder are regarding Gray value in Pin is more or less the same, even with means such as additional light source, can not accurately to slot size, The faults such as sieve plate fracture are monitored.

(3) feature analysis

Feature vector method is installation vibrating sensor on sieve plate, extracts the time domain of sieve plate, frequency domain, earthquake intensity etc. Feature, by sorting techniques such as BP neutral net, Bayesian Estimation, by normal for sieve plate and improper time spy Levy classification out, in order to determine whether current sieve plate is in the fault such as fracture, sieve aperture damage.The method can Realizing fault diagnosis to a certain extent, it is contemplated that the accuracy of method and stability, the method is not It is suitable for the production environment of coal preparation plant.

Although above method all carries out fault diagnosis with sieve plate for object, but they process in actual applications Cycle length, affected by environment compared with big, complexity is high, is not suitable for the specific environment that coal preparation plant produces.

Summary of the invention

The technical problem to be solved in the present invention is to provide a kind of dynamic sieve fault diagnosis analyzing model based on Weibull Method, solution is moved long processing period in the deficiency and actual coal separation production that in sieve monitoring, each method shows, is subject to The problems such as environmental effect is relatively big, complexity is high.

The technical scheme is that

A kind of dynamic sieve method for diagnosing faults analyzing model based on Weibull, this dynamic sieve method for diagnosing faults, passes through Shock sensors gathers the coal cinder of the driven sieve plate whereabouts impact dynamics to buffer board, calculates and is just setting out sieve plate Account for the specific gravity range of sample total amount, i.e. coal cinder particle size distribution through the coal cinder amount moving sieve plate time often；Time Between analyze the process that coal cinder falls in section t, total impulsive force P gathered according to shock sensors, when calculating Between in section t the total amount of dynamic sieve plate whereabouts coal cinder account for the proportion of sample total amount normal with dynamic sieve plate time specific gravity range enter Row contrast, thus measure whereabouts coal cinder total amount to dynamic sieve diagnosing malfunction:

Assuming that circular hole moves the aperture of sieve plate is d, and the parameter of coal crusher cutting method is designated as λ, according to dynamic sieve plate The shock sensors number arranged is n, in time period t, and the impulsive force accumulation of i-th sensor acquisition Amount is f_i, total impulsive force of all shock sensors collections is P, in the case of not having specified otherwise, above Symbolic unit all uses international basic unit:

Specifically comprising the following steps that of dynamic sieve method for diagnosing faults

(1), n shock sensors is installed on buffer board, measures when coal cinder falls and buffer board is rushed Hit dynamics；Set impulsive force lower limit as a；Record dynamic hole diameter of sieve (perforated) plate size d and parameter ξ of Rock and coal cutting method； Impact dynamics is periodically detected；

(2), coal cinder particle size distribution is calculated

First, calculate, according to Weibull distribution model, the coal cinder particle size distribution that current dynamic sieve plate falls, see formula (1):

W=1-exp (-ξ d^b) (1),

In above formula, b is breakage properties index, takes 0.4 1.3；Coal cinder particle size distribution W is also through diameter d The coal cinder amount of dynamic sieve plate accounting in sample total amount, is shown in formula (2):

In time period t, the summation of n shock sensors is:

$P=\underset{i=1}{\overset{n}{\Σ}}{f}_i---\left(3\right),$

Clash into the data analysis of buffer board according to coal cinder, the impulsive force of buffer board is shown in formula (4) by coal cinder:

$f=2.108\×m^{\frac{2}{3}}\·{\mathrm{\λ}}^{\frac{2}{5}}\·H^{\frac{3}{5}}---\left(4\right),$

In above formula, λ is Lame constants, is 10⁶N/m², H is the height of coal cinder freely falling body, is dynamic sieve The plate height away from buffer board, f is the coal cinder of the driven sieve plate whereabouts impulsive force to buffer board, and m is coal cinder quality；

The impulsive force that i-th shock sensors is measured is set to f_i, corresponding coal cinder quality is m_i, driven sieve Plate always whereabouts coal cinder quality is M_Fall, i.e. formula (5):

Formula (6) can be obtained by formula (3), (4) and (5),

The Weibull distribution W that coal cinder particle size distribution meets is fixed value, if the coal cinder dropping to dynamic sieve plate is total every time Mass M_AlwaysConstant, the most driven sieve plate drops to coal cinder gross mass M on buffer board_FallConstant, but actual survey During amount, the M recorded according to formula (6)_FallIt is change, due to M_AlwaysConstant, M_FallChanging value It is converted into Weibull distribution scope [W_L,W_H], i.e. M_FallWhen taking minima, then obtain W_L, M_FallTake maximum Time, then obtain W_H；

(3), dynamic sieve fault diagnosis:

Under conditions of Rock and coal cutting method and parameter constant, obtain coal cinder particle size distribution according to formula (1) full The value of the Weibull distribution W of foot is constant, but by formula (2), (6) calculated W value not at dynamic sieve plate Coal cinder particle size distribution [W after Rock and coal cutting time lossless_L,W_HIn], then can determine that dynamic sieve plate breaks down.

In described step (3) during dynamic sieve fault diagnosis, M_FallValue uses repetitive measurement to calculate multiple W value, If multiple W values are not the most at coal cinder particle size distribution [W_L,W_HIn], then can determine that dynamic sieve plate breaks down.

The invention has the beneficial effects as follows:

Due to the fact that and have employed such scheme, the method only drops to buffering by the analysis driven sieve plate of coal cinder The impulsive force of plate, is converted into coal cinder quality, uses coal cinder granularity to meet the theory of Weibull distribution model, The failure condition of dynamic sieve plate is analyzed, it is possible to overcome existing method for diagnosing faults cycle length, by environment because of The problems such as element impact is bigger, may be directly applied in the production of coal preparation plant.The abrasion change of existing sieve aperture is sifted out greatly when dynamic Or during compass screen surface fracture, the impulsive force of buffer board can be changed by the coal cinder fallen from sieve plate.Normal coal cinder Granularity less than sieve aperture, dynamic sieve does not has the granularity of coal cinder seriously beyond slot size when damaging, therefore to buffer board Impulsive force different, the relation of impulsive force Yu quality can calculate from the gross mass of sieve plate whereabouts coal cinder the most not Identical.It is dynamic with normal that dynamic sieve method for diagnosing faults based on coal cinder particle size distribution measures coal cinder particle size distribution exactly The coal cinder particle size distribution of sieve plate contrasts, and gathers the coal cinder impulsive force indirect measuring to buffer board by the cycle Go out coal cinder particle size distribution, and then to whether judging dynamic sieve fault.

Advantage: the present invention using coal cinder to the shock dynamics of buffer board as parameter, calculate coal cinder particle size distribution, And then the failure condition of the dynamic sieve of diagnosis, it is possible to more accurately, quickly, simply detect whether sieve plate occurs Fault.Due to can be by monitoring impulsive force in real time, the present invention can real time on-line monitoring dynamic sieve fault Situation, contributes to production and management personnel quickly carry out troubleshooting, reduces economic loss, and remembers for a long time The failure condition of the dynamic sieve of record can judge dynamic sieve inaction interval accurately, and the maintenance for dynamic sieve provides support. The present invention be a kind of efficiently, quick, simple, the dynamic sieve diagnostic method of low cost, automatic in intelligent mine Change production field and there is good application prospect.

Accompanying drawing explanation

Fig. 1 is the coal cinder of the present invention driven sieve plate dropping process figure to buffer board.

Detailed description of the invention

A kind of dynamic sieve method for diagnosing faults analyzing model based on Weibull, it is characterised in that: this dynamic sieve fault is examined Disconnected method, gathers the coal cinder of the driven sieve plate whereabouts impact dynamics to buffer board, meter by shock sensors Calculate dynamic sieve plate normal time account for the specific gravity range of sample total amount through the coal cinder amount of dynamic sieve plate, i.e. coal cinder granularity is divided Cloth scope；The process that coal cinder falls is analyzed, the total impact gathered according to shock sensors in time period t Power P, calculating the total amount of dynamic sieve plate whereabouts coal cinder in time period t, to account for the proportion of sample total amount normal with dynamic sieve plate Time specific gravity range contrast, thus measure whereabouts coal cinder total amount to dynamic sieve diagnosing malfunction:

Assuming that circular hole moves the aperture of sieve plate is d, and the parameter of coal crusher cutting method is designated as λ, according to dynamic sieve plate The shock sensors number arranged is n, in time period t, and the impulsive force accumulation of i-th sensor acquisition Amount is f_i, total impulsive force of all shock sensors collections is P, in the case of not having specified otherwise, above Symbolic unit all uses international basic unit:

Specifically comprising the following steps that of dynamic sieve method for diagnosing faults

(1), n shock sensors is installed on buffer board, measures when coal cinder falls and buffer board is rushed Hit dynamics；Set impulsive force lower limit as a；Record dynamic hole diameter of sieve (perforated) plate size d and parameter ξ of Rock and coal cutting method； Impact dynamics is periodically detected；

(3), coal cinder particle size distribution is calculated

First, calculate, according to Weibull distribution model, the coal cinder particle size distribution that current dynamic sieve plate falls, see formula (1):

W=1-exp (-ξ d^b) (1),

In above formula, b is breakage properties index, takes 0.4 1.3；Coal cinder particle size distribution W is also through diameter d The coal cinder amount of dynamic sieve plate accounting in sample total amount, is shown in formula (2):

In time period t, the summation of n shock sensors is:

$P=\underset{i=1}{\overset{n}{\Σ}}{f}_i---\left(3\right),$

Clash into the data analysis of buffer board according to coal cinder, the impulsive force of buffer board is shown in formula (4) by coal cinder:

$f=2.108\×m^{\frac{2}{3}}\·{\mathrm{\λ}}^{\frac{2}{5}}\·H^{\frac{3}{5}}---\left(4\right),$

In above formula, λ is Lame constants, is 10⁶N/m², H is the height of coal cinder freely falling body, is dynamic sieve The plate height away from buffer board, f is the coal cinder of the driven sieve plate whereabouts impulsive force to buffer board, and m is coal cinder quality；

The impulsive force that i-th shock sensors is measured is set to f_i, corresponding coal cinder quality is m_i, driven sieve Plate always whereabouts coal cinder quality is M_Fall, i.e. formula (5):

Formula (6) can be obtained by formula (3), (4) and (5),

The Weibull distribution W that coal cinder particle size distribution meets is fixed value, if the coal cinder dropping to dynamic sieve plate is total every time Mass M_AlwaysConstant, the most driven sieve plate drops to coal cinder gross mass M on buffer board_FallConstant, but actual survey During amount, the M recorded according to formula (6)_FallIt is change, due to M_AlwaysConstant, M_FallChanging value It is converted into Weibull distribution scope [W_L,W_H], i.e. M_FallWhen taking minima, then obtain W_L, M_FallTake maximum Time, then obtain W_H；

(3), dynamic sieve fault diagnosis:

Under conditions of Rock and coal cutting method and parameter constant, obtain what coal cinder particle size distribution met according to formula (1) The value of Weibull distribution W is constant, but is calculated multiple W value by formula (2), (6) repetitive measurement, if many The individual W value the most not coal cinder particle size distribution [W after Rock and coal cutting when dynamic sieve plate is lossless_L,W_HIn], then can determine that dynamic Sieve plate breaks down.

Claims (2)

1. the dynamic sieve method for diagnosing faults analyzing model based on Weibull, it is characterised in that: this dynamic sieve event Barrier diagnostic method, gathers the coal cinder of the driven sieve plate whereabouts impact dynamics to buffer board by shock sensors, Calculate and set out to account for the specific gravity range of sample total amount, i.e. coal cinder granularity through the coal cinder amount of dynamic sieve plate when sieve plate is normal Distribution；The process that coal cinder falls is analyzed, the total punching gathered according to shock sensors in time period t Hitting power P, the total amount moving sieve plate whereabouts coal cinder in calculating time period t with dynamic sieve plate is just accounting for the proportion of sample total amount Time often, specific gravity range contrasts, thus measures whereabouts coal cinder total amount to dynamic sieve diagnosing malfunction:

Assuming that circular hole moves the aperture of sieve plate is d, and the parameter of coal crusher cutting method is designated as λ, according to dynamic sieve plate The shock sensors number arranged is n, in time period t, and the impulsive force accumulation of i-th sensor acquisition Amount is f_i, total impulsive force of all shock sensors collections is P, in the case of not having specified otherwise, above Symbolic unit all uses international basic unit:

Specifically comprising the following steps that of dynamic sieve method for diagnosing faults

(1), n shock sensors is installed on buffer board, measures when coal cinder falls and buffer board is rushed Hit dynamics；Set impulsive force lower limit as a；Record dynamic hole diameter of sieve (perforated) plate size d and parameter ξ of Rock and coal cutting method； Impact dynamics is periodically detected；

(2), coal cinder particle size distribution is calculated

First, calculate, according to Weibull distribution model, the coal cinder particle size distribution that current dynamic sieve plate falls, see formula (1):

W=1-exp (-ξ d^b) (1),

In above formula, b is breakage properties index, takes 0.4 1.3；Coal cinder particle size distribution W is also through diameter d The coal cinder amount of dynamic sieve plate accounting in sample total amount, is shown in formula (2):

In time period t, the summation of n shock sensors is:

$P=\underset{i=1}{\overset{n}{\Σ}}{f}_i---\left(3\right),$

Clash into the data analysis of buffer board according to coal cinder, the impulsive force of buffer board is shown in formula (4) by coal cinder:

$F=2.108\×m^{\frac{2}{3}}\·{\mathrm{\λ}}^{\frac{2}{5}}\·H^{\frac{3}{5}}---\left(4\right),$

In above formula, λ is Lame constants, is 10⁶N/m², H is the height of coal cinder freely falling body, is dynamic sieve The plate height away from buffer board, f is the coal cinder of the driven sieve plate whereabouts impulsive force to buffer board, and m is coal cinder quality；

The impulsive force that i-th shock sensors is measured is set to f_i, corresponding coal cinder quality is m_i, driven sieve Plate always whereabouts coal cinder quality is M_Fall, i.e. formula (5):

Formula (6) can be obtained by formula (3), (4) and (5),

The Weibull distribution W that coal cinder particle size distribution meets is fixed value, if the coal cinder dropping to dynamic sieve plate is total every time Mass M_AlwaysConstant, the most driven sieve plate drops to coal cinder gross mass M on buffer board_FallConstant, but actual survey During amount, the M recorded according to formula (6)_FallIt is change, due to M_AlwaysConstant, M_FallChanging value It is converted into Weibull distribution scope [W_L,W_H], i.e. M_FallWhen taking minima, then obtain W_L, M_FallTake maximum Time, then obtain W_H；

(3), dynamic sieve fault diagnosis:

Under conditions of Rock and coal cutting method and parameter constant, obtain coal cinder particle size distribution according to formula (1) full The value of the Weibull distribution W of foot is constant, but by formula (2), (6) calculated W value not at dynamic sieve plate Coal cinder particle size distribution [W after Rock and coal cutting time lossless_L,W_HIn], then can determine that dynamic sieve plate breaks down.

A kind of dynamic sieve method for diagnosing faults analyzing model based on Weibull the most according to claim 1, It is characterized in that: in described step (3) during dynamic sieve fault diagnosis, M_FallValue uses repetitive measurement to calculate Multiple W values, if multiple W value is not the most at coal cinder particle size distribution [W_L,W_HIn], then can determine that dynamic sieve plate occurs Fault.