CN106292612A - A kind of ladle baking facility on-line fault diagnosis system - Google Patents

Abstract

The invention provides a kind of ladle baking facility on-line fault diagnosis system, including data acquisition transmission system, data server and central processing unit；Described data acquisition transmission system, for gathering the process data of ladle baking facility and process data being sent to data server；Described data server is for being uploaded to described central processing unit by the process data received；Described central processing unit, for the process data of the ladle baking facility received being stored in the problem case storehouse of described central processing unit, data in the case library with described central processing unit of every the problem data in problem case storehouse are carried out similarity comparison, using fault type corresponding for the data in case library maximum for similarity as the fault diagnosis result of this problem data, and described fault diagnosis result is pushed out.The present invention can improve reliability and the efficiency of diagnostic result, it is achieved the on-line monitoring of ladle baking facility, it is ensured that safety in production.

Description

A kind of ladle baking facility on-line fault diagnosis system

Technical field

The invention monitoring technical field, is specifically related to a kind of ladle baking facility on-line fault diagnosis system.

Background technology

Owing to ladle baking facility is more at each steel mill's station and it is very frequent to use, event occurs during use unavoidably Hinder, and fault has the strongest hysteresis quality.In correlation technique, only after field worker discovering device or data exception, just meeting Arrangement personnel keep in repair ladle baking facility, and this maintenance mode afterwards can affect production run, thereby increases and it is possible to can cause peace Full accident.Owing to steel mill ladle baking facility quantity is more, and need to be acquired its process data in real time processing, and pass through Prediction judges whether this ladle baking facility breaks down, single or minicomputer treatment effeciency is low, and can not be real-time right Ladle baking facility carries out real-time monitoring.

Summary of the invention

The purpose of the invention is to solve above-mentioned weak point of the prior art and provides a kind of ladle baking facility On-line fault diagnosis system.

The purpose of the invention is achieved through the following technical solutions:

A kind of ladle baking facility on-line fault diagnosis system, including data acquisition transmission system, data server and central authorities Processor；Described data acquisition transmission system, for gathering the process data of ladle baking facility and process data being sent to number According to server；Described data server is for being uploaded to described central processing unit by the process data received；Described centre Reason device, for the process data of the ladle baking facility received is stored in the problem case storehouse of described central processing unit, will Every problem data in problem case storehouse carries out similarity comparison with the data in the case library of described central processing unit, Using fault type corresponding for the data in case library maximum for similarity as the fault diagnosis result of this problem data, And described fault diagnosis result is pushed out.

The beneficial effect of the invention: reliability and the efficiency of diagnostic result can be improved, and push fault diagnosis knot Fruit reminds staff in advance the ladle baking facility that will break down to be overhauled and adjusted corresponding steel-making plan arrangement, Realize the on-line monitoring of ladle baking facility, it is ensured that safety in production.

Accompanying drawing explanation

Utilize accompanying drawing that innovation and creation are described further, but the embodiment in accompanying drawing does not constitute and appoints the invention What limits, for those of ordinary skill in the art, on the premise of not paying creative work, it is also possible to according to the following drawings Obtain other accompanying drawing.

Fig. 1 is present configuration connection diagram；

Fig. 2 is the structural representation of data acquisition transmission system.

Reference:

Data acquisition transmission system 1, data server 2, central processing unit 3, mobile terminal 4, data monitoring unit 11, number According to processing unit 12, data transmission unit 13.

Detailed description of the invention

The invention will be further described with the following Examples.

Seeing Fig. 1, Fig. 2, the ladle baking facility on-line fault diagnosis system of the present embodiment, including data acquisition transmission system 1, data server 2 and central processing unit 3；Described data acquisition transmission system 1, for gathering the process data of ladle baking facility And process data is sent to data server 2；Described data server 2 is described for the process data received being uploaded to Central processing unit 3；Described central processing unit 3, for being stored in described central authorities by the process data of the ladle baking facility received In the problem case storehouse of processor 3, by the history case of every problem data in problem case storehouse Yu described central processing unit 3 Data in storehouse carry out similarity comparison, using fault type corresponding for the data in the case library of similarity maximum as this The fault diagnosis result of bar problem data, and described fault diagnosis result is pushed out.

Preferably, described system also includes that mobile terminal 4, described mobile terminal 4 are connected with described central processing unit 3, uses In receiving the fault diagnosis result that described central processing unit 3 pushes.

Preferably, described mobile terminal 4 is mobile phone.

The above embodiment of the present invention can improve reliability and the efficiency of diagnostic result, and pushes fault diagnosis result prompting The ladle baking facility that will break down is overhauled and adjusts corresponding steel-making plan arrangement by staff in advance, it is achieved steel The on-line monitoring of bag roaster, it is ensured that safety in production.

Preferably, described data acquisition transmission system 1 includes that data monitoring unit 11, data processing unit 12 and data pass Defeated unit 13；Described data monitoring unit 11 carries out ladle for being cooperated by the sensor node built by each sensor The monitoring of roaster, and export the process data of the ladle baking facility of each sensor node monitoring；Described data processing unit 12 Process data for monitoring each sensor node carries out pretreatment, to obtain effective process data；Described data are transmitted Unit 13 carries out the transmission of process data for using data transmission mechanism set in advance.

Preferably, described data monitoring unit 11 includes sensor locator unit and data correction subelement；Described biography Sensor locator unit is used for carrying out sensor node localization, described in carry out sensor node localization and specifically include:

(1) by small part sensor node deployment on known position coordinates, as beaconing nodes, according to beaconing nodes The local coordinate system meeting domestic environment is set up in position；

(2) assume the coordinate of unknown node for (x, y), it is possible to the beaconing nodes coordinate being in communication with is respectively (x₁, y₁), (x₂,y₂) ..., (x_n,y_n), it is the center of circle that unknown node is positioned at each beaconing nodes, and communication radius is the intersection of the circle of radius Territory, takes this region any point as the pre-reconnaissance of unknown node coordinate, the difference letter of the pre-reconnaissance of unknown node coordinate to beaconing nodes Number is:

$d(x,y)={\mathrm{\Σ}}_{i=1}^n|\sqrt{{(x-x_i)}^2+{(x-x_i)}^2}-r_i|$

In formula, r_iFor the pre-reconnaissance of unknown node coordinate to (x_i,y_i) distance, i=1,2 ..., n, utilize maximal possibility estimation Method asks for the minima of this formula, is unknown node coordinate position；

Described data correction subelement for sensor Monitoring Data under non-standard environmental is modified, modifying factor Sub-σ is:

$\mathrm{\&sigma;}=\left\{\begin{array}{c}{e}^{-\sqrt{\frac{{(T-T_0)}^2}{T^2}}},T\&GreaterEqual;T_0\\ e^{\sqrt{\frac{T^2}{{(T-T_0)}^2}}},T<T_0\end{array}\right.$

In formula, T₀The standard temperature used for sensor, T is ambient temperature during sensor use.

This preferred embodiment achieves the accurate measurement of sensor location and data.

Preferably, described data processing unit 12 includes data filtering module and Supplementing Data module, described data filtering Module is for by the process data of intelligent gateway filter false, and described Supplementing Data module is for filling up the number of passes excessively of loss According to；

The described process data by intelligent gateway filter false, including:

(1) process data in the effective range not falling within sensor reading is removed；

(2) remaining process data is carried out filtration treatment, particularly as follows:

1) assume that the sensing frequency of all the sensors is consistent and sliding window size is A, utilize Jaccard similarity function Calculate two sensor S_I、S_JSimilarity Sim (the x of reading behavior_I(t),x_J(t)), wherein x_IT () represents sensor S_ITime Between t time level readings, x_JT () represents sensor S_JLevel readings when time t；

2) definition sensor S_I、S_JInitially interim similarity be:

${\mathrm{Sim}}_L^0(I,J)=Sim\left(x_I\right(A+t),x_J(A+t\left)\right)$

The interim similarity in K+1 moment is:

${\mathrm{Sim}}_L^{K+1}(I,J)={\mathrm{\&mu;Sim}}_L^K(I,J)+\frac{1+\mathrm{\&mu;}}{2}Sim\left(x_I\right(t),x_J(t\left)\right)$

In formula, x_I(A+t) sensor S is represented_IInitial reading behavior, μ is the adjustable variable set, the value model of μ Enclose for [-1,1]；

3) total M sensor is set, from sensor S_IAn event query q, frequently inquired about by sensor in the past Data than those seldom by sensor S_IThe data read and sensor S_IHave more dependency, calculate sensor S_ISequence Grade point Q_L(q,S_I), computing formula is:

$Q_L(q,S_I)=\frac{\underset{i\&Element;\&lsqb;1,M\&rsqb;}{max}{\mathrm{Sim}}_L^{K+1}(I,J)}{{\mathrm{\&Sigma;}}_{J=1}^M{\mathrm{Sim}}_L^{K+1}(I,J)}\&times;\frac{\left|n(q,S_I)\right|}{\left|N(q,S_I)\right|+B}$

In formula, n (q, S_I) it is at past sensor S_IFor the inquiry q event query number to some themes, N (q, S_I) Represent sensor S_IFor the event query sum of inquiry q, B is smoothing factor；

4) ballot method is used to judge sensor S_ICurrent reading whether be false readings, set discriminant function as:

$P_I=\underset{S_J\&Element;net\left(I\right)}{\&Sigma;}{\mathrm{VDecision}}_J\left(I\right)\&CenterDot;Q_L(q,S_I)$

In formula, net (I) represents sensor S_INeighbours' set of sensors, VDecision_j(I) neighbours sensor S is represented_J To sensor S_IMaking and choosing in a vote, chosen two classes in a vote, a class is positive, and another kind of is passive, choose in a vote into Time actively, VDecision_j(I)=1, when choosing in a vote as passiveness, VDecision_j(I)=0；P_I> 0 time, represent sensor S_I Current reading be normal reading, P_I< when 0, represent sensor S_ICurrent reading be false readings.

The described process data filling up loss, including:

(1) K-means clustering method is used to carry out the clustering processing of process data；

(2) to the missing values in same class but it is the filling of data.

During the process data of this preferred embodiment filter false, it is contemplated that between sensor and neighbor node and environment Relatedness, is the importance that each sensor distributes that a suitable weight reflects sensor by sort algorithm, and then right Sensor acquisition to the abnormal data that data carry out accepting or rejecting the wrong data filtering in gatherer process and environment causes, carry The high precision filtered；Use first to cluster and fill up the mode of missing values afterwards and fill up the process data of loss, can preferably consider The local characteristics of data, improves the precision of data filling.

In one embodiment, described data transmission mechanism set in advance includes:

Bunch (1) in member within the time period of each distribution, the data collected are transferred in the way of single-hop belonging to bunch Leader cluster node C_i, leader cluster node C_iThe process data collected is carried out integration process；

(2) leader cluster node C_iOther leader cluster nodes are selected, to determine its route candidate cluster head node set CN according to following formula:

$CN=\left\{\begin{array}{c}{d}_{jB}<d_{iB}\\ d_{ij}<d_{iB}\\ {\mathrm{NS}}_i<{\mathrm{NS}}_j\end{array}\right\}$

In formula, d_iBRepresent leader cluster node C_iTo the distance of base station, d_jBRepresent leader cluster node C_jTo the distance of base station, d_ijRepresent Leader cluster node C_iTo leader cluster node C_jDistance, NS_iRepresent leader cluster node C_iDump energy rank, NS_jRepresent leader cluster node C_j Dump energy rank；

(3) if route candidate cluster head node set CN is empty, then leader cluster node C_iDirectly process data is sent to base station, And jump to (6)；

(4) if route candidate cluster head node set only exists a ratio leader cluster node C_iDistance base station closer to or residue Energy rank other leader cluster nodes higher C_j, then leader cluster node C is selected_jAs down hop routing node, and process data is turned Issue leader cluster node C_j, make leader cluster node C_jBecome new leader cluster node C_i, and jump back to (2)；

(5) if route candidate cluster head node set exists multiple ratio leader cluster node C_iDistance base station closer to or residual energy Magnitude other leader cluster nodes the most higher C_j, then the leader cluster node C making repeating process data communication expense minimum is selected_jAs Down hop routing node, and process data is transmitted to leader cluster node C_j, make leader cluster node C_jBecome new leader cluster node C_i, and Jump back to (2)；

(6) base station receives leader cluster node C_iThe process data sent.

This preferred embodiment uses data transmission mechanism set in advance to carry out the transmission of process data, makes leader cluster node Dump energy distribution more equalizes, and relatively closely or distance base station node energy farther out consumes too early to efficiently solve distance base station Complete problem, thus extend the life cycle of whole data transmission network.

In another embodiment, described data transmission mechanism set in advance includes:

Transmit size of data for difference and use different transmission means with transmission range, definition transfer function Tr:

$Tr=\frac{S}{S_0}+\frac{D}{D_0}$

In formula, S is transmission size of data, S₀For data transfer size cut off value, S₀=1MB, D are transmission range, D₀For passing Defeated distance cut off value, D₀=10m；

If D is < D₀, then use bluetooth to carry out data communication,

If D >=D₀And S≤S₀, then use zigbee network to carry out data communication,

If D >=D₀And S > S₀, then WIFI is used to communicate.

This preferred embodiment achieves the communication of the low energy consumption of the process data of ladle baking facility, two-forty.

Last it should be noted that, above example is only in order to illustrate technical scheme, rather than the present invention is protected Protecting the restriction of scope, although having made to explain to the present invention with reference to preferred embodiment, those of ordinary skill in the art should Work as understanding, technical scheme can be modified or equivalent, without deviating from the reality of technical solution of the present invention Matter and scope.

Claims (3)

1. a ladle baking facility on-line fault diagnosis system, is characterized in that, including data acquisition transmission system, data server And central processing unit；Described data acquisition transmission system, for gathering the process data of ladle baking facility and process data being passed Deliver to data server；Described data server is for being uploaded to described central processing unit by the process data received；Described Central processing unit, for being stored in the problem case storehouse of described central processing unit by the process data of the ladle baking facility received In, the data in the case library with described central processing unit of every the problem data in problem case storehouse are carried out similarity Contrast, using fault type corresponding for the data in case library maximum for similarity as the fault diagnosis of this problem data As a result, and by described fault diagnosis result push out.

A kind of ladle baking facility on-line fault diagnosis system the most according to claim 1, is characterized in that, described system is also wrapped Including mobile terminal, described mobile terminal is connected with described central processing unit, for receiving the fault that described central processing unit pushes Diagnostic result.

A kind of ladle baking facility on-line fault diagnosis system the most according to claim 2, is characterized in that, described mobile terminal For mobile phone.