CN117382967A

CN117382967A - Method, device and equipment for determining running state of cigarette packing machine

Info

Publication number: CN117382967A
Application number: CN202311363297.3A
Authority: CN
Inventors: 刘钊; 吴爱民; 吴涛; 郭雪刚; 马秀伟; 赵立辉; 高欣; 李练兵
Original assignee: Hebei Baisha Tobacco Co Ltd
Current assignee: Hebei Baisha Tobacco Co Ltd
Priority date: 2023-10-19
Filing date: 2023-10-19
Publication date: 2024-01-12

Abstract

The invention provides a method, a device and equipment for determining the running state of a cigarette packing machine, wherein the method comprises the following steps: acquiring an operation data set of the target cigarette packaging machine in the current period, and inputting the operation data set into a trained fault state determination model to determine the operation state of the target cigarette packaging machine; the fault state determining model is obtained based on a first decision hyperplane and a second decision hyperplane, the first decision hyperplane is constructed based on a first training sample set formed by basic control parameter sets of all training samples, the second decision hyperplane is constructed based on a second training sample set formed by preset control parameter sets of all training samples, the basic control parameter sets are extracted from the preset control parameter sets according to preset extraction rules, and the running state comprises a normal state, a disturbance state and a fault state. The method can improve the accuracy of the running state prediction of the cigarette packer.

Description

Method, device and equipment for determining running state of cigarette packing machine

Technical Field

The invention relates to the technical field of detection of tobacco mechanical equipment, in particular to a method, a device and equipment for determining the running state of a cigarette packaging machine.

Background

Along with the progress and development of technology, the automation level of industrial production is also higher and higher. Cigarette enterprises are also increasingly dedicated to realizing the 'careless production' of cigarettes and the near zero fault of cigarette packaging machine equipment and the high-reliability production of packaging machines.

The fault detection and diagnosis of cigarette packer equipment has undergone a development process from downtime maintenance, periodic maintenance to predictive maintenance. Early downtime maintenance does not truly circumvent the economic loss. Periodic maintenance or computational maintenance can be excessive and cause unnecessary resource waste.

At present, although a related prediction method for predictive maintenance of the faults of the cigarette packing machine also exists, the accuracy of prediction is poor, the pre-maintenance and the normal maintenance of the faults of the packing machine cannot be realized, and the loss caused by the untimely fault clearing of the cigarette packing machine is increased.

Disclosure of Invention

The embodiment of the invention provides a method, a device and equipment for determining the running state of a cigarette packing machine, which are used for solving the problem that the running state of the packing machine cannot be accurately determined at present.

In a first aspect, an embodiment of the present invention provides a method for determining an operating state of a cigarette packer, including:

acquiring a target operation data set of a target cigarette packaging machine in a current period, wherein the target operation data set is a data set obtained by detecting a preset control parameter set in real time when the target cigarette packaging machine operates, and the preset control parameter set comprises a plurality of control parameters;

Inputting the target operation data set into a trained fault state determination model to determine the operation state of the target cigarette packaging machine; the fault state determining model is obtained based on a first decision hyperplane and a second decision hyperplane, the first decision hyperplane is constructed based on a first training sample set formed by basic control parameter sets of all training samples, the second decision hyperplane is constructed based on a second training sample set formed by preset control parameter sets of all training samples, the basic control parameter sets are extracted from the preset control parameter sets according to preset extraction rules, the training samples comprise an operation data set in normal operation, an operation data set in disturbance state and an operation data set in fault state, each data set is provided with a corresponding data set state label, and the operation states comprise a normal state, a disturbance state and a fault state.

In one possible implementation, the method further includes:

when the target cigarette packaging machine is determined to be in a fault state, sending a feedback signal of the fault of the cigarette packaging machine;

comparing the standardized target operation data set with a standard data set, and determining a significant control parameter set deviating from a normal working state, wherein the significant control parameter set at least comprises one control parameter;

And searching based on the significant control parameter set and a pre-constructed fault tree, and determining the fault type corresponding to the fault state.

In one possible implementation, the standard dataset includes standard values for each control parameter;

comparing the standardized target operation data set with the standard data set, and determining a significant control parameter set deviating from the normal working state, wherein the method comprises the following steps:

determining the deviation degree of the target control parameter based on the actual value of the target control parameter and the standard value of the target control parameter, wherein the target control parameter is any one control parameter;

determining a significant control parameter set deviating from a normal working state based on the degree of deviation of all the target control parameters;

degree of deviation theta _i The calculation mode of (a) is as follows:

A _i is the actual value of the i-th control parameter,is the standard value of the ith control parameter.

In one possible implementation, the method further includes:

when the target cigarette packing machine is determined to be in a disturbance state, sending a feedback signal of disturbance of the cigarette packing machine;

inputting the target operation data set and a time sequence corresponding to the target operation data set into a trained time sequence prediction model, and predicting the operation data set of the target cigarette packaging machine in a preset period, wherein the preset period is a period after the current period;

And inputting the operation data set in a preset period into a fault state determination model to determine the operation state trend of the target cigarette packaging machine.

In one possible implementation, the basic set of control parameters is determined based on correlation coefficients between all control parameters in the preset set of control parameters and a preset correlation coefficient threshold;

the preset control parameter set comprises the number of cigarettes lacking of a hopper branch, the number of damaged cigarettes in a cigarette delivering mould box, the number of cigarettes continuously removed in the cigarette delivering mould box, the number of aluminum foil papers, the position of the aluminum foil papers, the length of the aluminum foil papers, the rotating speed of an aluminum foil cigarette packet feeding wheel, the rotating speed of an aluminum foil cigarette packet discharging wheel, the content of inner frame papers, the content of trademark papers, the rotating speed of a trademark paper feeding wheel and the rotating speed of a trademark paper discharging wheel;

the basic control parameter set comprises the number of cigarettes lacking of the hopper branches, the number of damaged cigarettes in the cigarette feeding mould box, the number of continuous cigarettes removed in the cigarette feeding mould box, the number of aluminum foil papers, the position of the aluminum foil papers, the rotating speed of the aluminum foil cigarette packet feeding wheel, the content of the inner frame papers, the content of the label papers and the rotating speed of the label paper feeding wheel.

In one possible implementation manner, the fault state determining model includes 3 decision hyperplanes, wherein the first decision hyperplane is obtained by weighting a normal fault first decision hyperplane constructed based on a normal state in the training sample and a corresponding basic control parameter set under the fault state and a normal fault second decision hyperplane constructed based on a normal state in the training sample and a corresponding preset control parameter set under the fault state; the second decision hyperplane is obtained by weighting a normal disturbance first decision hyperplane constructed based on a corresponding basic control parameter set in a normal state and a disturbance state in the training sample and a normal disturbance second decision hyperplane constructed based on a corresponding preset control parameter set in the normal state and the disturbance state in the training sample; the third decision hyperplane is obtained by weighting a first decision hyperplane of fault disturbance constructed based on a corresponding basic control parameter set in a fault state and a disturbance state in a training sample and a second decision hyperplane of fault disturbance constructed based on a corresponding preset control parameter set in the fault state and the disturbance state in the training sample;

Inputting the target operating dataset into a trained fault state determination model to determine an operating state of the target cigarette packaging machine, comprising:

inputting the target operation data set into the first decision hyperplane, the second decision hyperplane and the third decision hyperplane respectively to obtain all the output classification results;

and determining the most classification result in all classification results as the running state of the target cigarette packaging machine.

In a second aspect, an embodiment of the present invention provides a device for determining an operational state of a cigarette packer, including:

the acquisition module is used for acquiring a target operation data set of the target cigarette packaging machine in the current period, wherein the target operation data set is a data set obtained by detecting a preset control parameter set in real time when the target cigarette packaging machine operates, and the preset control parameter set comprises a plurality of control parameters;

the determining module is used for inputting the target operation data set into the trained fault state determining model so as to determine the operation state of the target cigarette packaging machine; the fault state determining model is obtained based on a first decision hyperplane and a second decision hyperplane, the first decision hyperplane is constructed based on a first training sample set formed by basic control parameter sets of all training samples, the second decision hyperplane is constructed based on a second training sample set formed by preset control parameter sets of all training samples, the basic control parameter sets are extracted from the preset control parameter sets according to preset extraction rules, the training samples comprise an operation data set in normal operation, an operation data set in disturbance state and an operation data set in fault state, each data set is provided with a corresponding data set state label, and the operation states comprise a normal state, a disturbance state and a fault state.

In one possible implementation manner, the determining module is used for sending a feedback signal of the fault of the cigarette packaging machine when the target cigarette packaging machine is determined to be in the fault state;

the determining module is used for determining the deviation degree of the target control parameter based on the actual value of the target control parameter and the standard value of the target control parameter, wherein the target control parameter is any one control parameter;

degree of deviation theta _i The calculation mode of (a) is as follows:

A _i is the average value of the i-th control parameter,is the standard value of the ith control parameter.

In one possible implementation, the standard value of the target control parameter is determined based on an average value of the target control parameter in all the normal operation data sets in the training sample.

In one possible implementation manner, the determining module is used for sending a feedback signal of disturbance of the cigarette packaging machine when the target cigarette packaging machine is determined to be in a disturbance state;

The basic control parameter set comprises the number of cigarettes lacking of the hopper branches, the number of damaged cigarettes in the cigarette feeding mould box, the number of continuous removing cigarettes in the cigarette feeding mould box, the number of aluminum foil papers, the position of the aluminum foil papers, the rotating speed of the aluminum foil cigarette packet feeding wheel, the content of the inner frame paper, the content of the label paper and the rotating speed of the label paper feeding wheel.

The determining module is used for respectively inputting the target operation data set into the first decision hyperplane, the second decision hyperplane and the third decision hyperplane to obtain all the output classification results;

In a third aspect, an embodiment of the present invention provides an electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the method according to the first aspect or any one of the possible implementations of the first aspect, when the computer program is executed by the processor.

In a fourth aspect, embodiments of the present invention provide a computer readable storage medium storing a computer program which, when executed by a processor, implements the steps of the method as described above in the first aspect or any one of the possible implementations of the first aspect.

The embodiment of the invention provides a method, a device and equipment for determining the running state of a cigarette packing machine. According to the invention, two decision hyperplanes are constructed by adopting a support vector machine, one is based on all control parameters, the other is based on basic control parameters, and finally, the two decision hyperplanes are weighted, so that the prediction accuracy of the fault state determination model is improved. The method for determining the running state of the cigarette packing machine can determine the current running state of the cigarette packing machine in time and accurately according to the running data set of the current time period, can greatly reduce the loss caused by untimely fault clearing of the cigarette packing machine, and is easy to realize the pre-maintenance and normal operation and maintenance of the cigarette packing machine.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments or the description of the prior art will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is a flowchart of an implementation of a method for determining an operating state of a cigarette packer according to an embodiment of the present invention;

fig. 2 is a schematic structural view of a determining device for an operating state of a cigarette packer according to an embodiment of the present invention;

fig. 3 is a schematic diagram of an electronic device according to an embodiment of the present invention.

Detailed Description

In the following description, for purposes of explanation and not limitation, specific details are set forth such as the particular system architecture, techniques, etc., in order to provide a thorough understanding of the embodiments of the present invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present invention with unnecessary detail.

For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the following description will be made by way of specific embodiments with reference to the accompanying drawings.

Along with the continuous development of technology, based on the existing numerical control products and the information physical system technology, the intelligent packaging machine can realize comprehensive monitoring on the running state of equipment, the technological state of the products and the external environment by adopting the means of the Internet of things and through the sensors and the external data acquisition, and realize the data acquisition and storage of the full state of the cigarette packaging machine, so that the cigarette packaging machine equipment has comprehensive sensing capability, state analysis and fault early warning capability, and the packaging machine can have the full life cycle knowledge base construction capability.

Therefore, a packing machine running state determining method with higher accuracy can be constructed based on the full life cycle knowledge base of the cigarette packing machine, so that the problem that the existing fault prediction accuracy is low and the faults cannot be predicted and found in time is solved.

In order to solve the problems in the prior art, the embodiment of the invention provides a method, a device, equipment and a storage medium for determining the running state of a cigarette packing machine. The following first describes a method for determining the operation state of the packaging machine according to the embodiment of the present invention.

Referring to fig. 1, a flowchart of an implementation of a method for determining an operating state of a cigarette packaging machine according to an embodiment of the present invention is shown, and details are as follows:

step S110, a target operation data set of the target cigarette packaging machine in the current period is obtained.

The target operation data set is a data set obtained by detecting a preset control parameter set in real time when the target cigarette packaging machine operates. The preset control parameter set includes a plurality of control parameters.

The operation data of the cigarette packing machine can embody the operation condition of each monitoring part of the packing machine, and whether the operation of each monitoring part is normal or not can be embodied according to each operation data.

The existing domestic tobacco packaging equipment mostly comprises a hard box packaging machine, an external transparent paper packaging machine, a hard strip packaging machine and the like, and when faults occur, the fault reasons are difficult to find rapidly by manpower only because the packaging machine comprises more parts, and the fault occurring positions are found timely according to the faults.

In some embodiments, the preset control parameter set of the packaging machine may include a missing number of cigarettes in the hopper branches, a damaged number of cigarettes in the cigarette delivery module, a uniformity of cigarettes in the cigarette delivery module, a number of aluminum foil papers, a position of the aluminum foil papers, a length of the aluminum foil papers, an in-wheel rotation speed of the aluminum foil cigarettes, an out-wheel rotation speed of the aluminum foil cigarettes, an in-frame paper content, a trademark paper content, an in-wheel rotation speed of the trademark paper, and an out-wheel rotation speed of the trademark paper.

The number of the cigarettes missing of the hopper branches can reflect whether the cigarettes are smooth or not, and whether continuous cigarette missing phenomenon occurs in the cigarette branching process or not. When the number of times of continuous smoke shortage of the channel in the hopper exceeds a set value, the operation is considered to be influenced, and shutdown maintenance is needed.

The number of damaged cigarettes in the cigarette feeding mould box can reflect whether the cigarettes are empty, damaged and the like, and can be detected by means of optical waveforms and the like.

The continuous removing quantity of cigarettes in the cigarette feeding die box can reflect whether the arrangement of cigarettes is irregular or not, and if the cigarettes exist, removing is carried out; when the number of continuous rejects exceeds a set value, the operating equipment needs to be stopped for inspection.

The number of the aluminum foil papers can reflect whether the stock quantity of the aluminum foil papers is lower than a set value, and when the stock quantity of the aluminum foil papers is lower than the set value, the paper feeding operation is needed.

The position of the aluminum foil paper can reflect whether the aluminum foil paper has offset or not.

The length of the aluminum foil paper can judge whether the aluminum foil paper breaks.

The aluminum foil cigarette packet wheel feeding rotating speed is used for being responsible for carrying out the rotating speed of the packaging of the aluminum foil cigarette packet, the aluminum foil cigarette packet wheel discharging rotating speed is used for being responsible for rejecting unqualified cigarette packets, and the aluminum foil cigarette packet wheel feeding rotating speed can be used for judging whether the aluminum foil cigarette packet is blocked.

The inner frame paper content can reflect whether the inner frame paper stock is lower than a set value, and when the inner frame paper stock is lower than the set value, paper feeding operation is needed.

The content of the label paper can reflect the amount of the label paper, and when the content is lower than a set value, paper adding operation is needed.

The label paper wheel feeding rotating speed is used for judging the folding speed of label paper, the label paper wheel discharging rotating speed is used for judging the conveying speed of the cigarette packet, and the label paper wheel feeding rotating speed can be used for judging whether the label paper is blocked.

And step S120, inputting the target operation data set into the trained fault state determination model to determine the operation state of the target cigarette packaging machine.

The fault state determination model is obtained based on a first decision hyperplane and a second decision hyperplane, wherein the first decision hyperplane is constructed based on a first training sample set formed by basic control parameter sets of training samples, and the second decision hyperplane is constructed based on a second training sample set formed by preset control parameter sets of the training samples. The basic control parameter set is extracted from the preset control parameter set according to a preset extraction rule.

In some embodiments, the fault state determination model includes 3 decision hyperplanes, where a first decision hyperplane is weighted by a first decision hyperplane constructed based on a normal state in the training sample and a corresponding basic control parameter set under the fault state and a normal fault second decision hyperplane constructed based on a normal fault normal state in the training sample and a corresponding preset control parameter set under the fault state. The second decision hyperplane is obtained by weighting a normal disturbance first decision hyperplane constructed based on a corresponding basic control parameter set in a normal state and a disturbance state in the training sample and a normal disturbance second decision hyperplane constructed based on a corresponding preset control parameter set in the normal state and the disturbance state in the training sample. The third decision hyperplane is obtained by weighting a first decision hyperplane of fault disturbance constructed based on a corresponding basic control parameter set in a fault state and a disturbance state in the training sample and a second decision hyperplane of fault disturbance constructed based on a corresponding preset control parameter set in the fault state and the disturbance state in the training sample.

For ease of understanding, the construction and training process of the fault state determination model will first be described herein.

The operation state of the cigarette packing machine is divided into a normal state and an abnormal state, and the abnormal state includes a disturbance state and a fault state. The disturbance state means that the operation state of the packaging machine is affected by small disturbance, so that the operation speed, the production quality and the like of the packaging machine are slightly reduced, but the operation of the packaging machine cannot be failed or even stopped. A fault condition means that the operation of the packaging machine is severely affected, possibly producing waste products and even a shutdown.

The target cigarette packaging machine is provided with an on-line monitoring system, and the running data of the target cigarette packaging machine at any moment can be acquired through the on-line monitoring system. In addition, when the cigarette packing machine works and has a disturbance state or a fault state, the staff can also make corresponding records. The staff can record detailed processing information such as occurrence time, reasons, solutions and the like of the disturbance state or the fault state.

The training samples of the fault state determination model comprise a normal operation data set, a disturbance state operation data set and a fault state operation data set which are extracted from the historical data of the target cigarette packaging machine. The label corresponding to the data in the operation data set in normal operation is 1, the label corresponding to the operation data set in disturbance state is 2, the label corresponding to the operation data set in fault state is 3, and the label is used for distinguishing the data in different operation states and preventing the confusion of the subsequent data processing. The operation data set in the disturbance state and the operation data set in the fault state are the data of the time period corresponding to the corresponding record of the target cigarette packer in the disturbance state and the fault state.

In some embodiments, after the training samples are obtained, whether the data is wrong or not is also required to be detected, and the data in the normal running data set, the running data set in the disturbance state and the running data set in the fault state in all the training samples are all data processed through at least one of discrete data rejection, difference value, fitting or standardization processing.

In this embodiment, a scatter plot of the 3 operational conditions of the normal state, the disturbance state, and the fault state may be plotted, respectively, such as may be provided using the t-sne algorithm and a scatter plot tool. When the positions of the scattered points are obviously far away from most data, the data of the corresponding time points can be deleted, and interpolation or fitting is needed after the data are deleted. After processing the data in the training samples, all the data also needs to be normalized. After the data is normalized, the training samples are divided into a training sample set and a test sample set.

The fault state determination model is a decision hyperplane constructed based on a support vector machine model. The support vector machine is a model algorithm designed for the two-classification problem, so that the three-classification problem of the three state categories of the normal state, the disturbance state and the fault state is split into a plurality of two-classification problems.

The first decision hyperplane is constructed based on a first training sample set consisting of basic control parameter sets of all training samples, and the second decision hyperplane is constructed based on a second training sample set consisting of preset control parameter sets of all training samples. The fault state determining model is obtained based on the first decision hyperplane and the second decision hyperplane through weighting processing, and the classification precision of the finally obtained decision hyperplane can be improved.

The fault state determining model comprises 3 decision hyperplanes, wherein the first decision hyperplane is obtained by weighting a normal fault first decision hyperplane constructed based on a normal state in a training sample and a corresponding basic control parameter set under the fault state and a normal fault second decision hyperplane constructed based on a normal state in the training sample and a corresponding preset control parameter set under the fault state. The second decision hyperplane is obtained by weighting a normal disturbance first decision hyperplane constructed based on a corresponding basic control parameter set in a normal state and a disturbance state in the training sample and a normal disturbance second decision hyperplane constructed based on a corresponding preset control parameter set in the normal state and the disturbance state in the training sample. The third decision hyperplane is obtained by weighting a first decision hyperplane of fault disturbance constructed based on a corresponding basic control parameter set in a fault state and a disturbance state in the training sample and a second decision hyperplane of fault disturbance constructed based on a corresponding preset control parameter set in the fault state and the disturbance state in the training sample.

According to the invention, the support vector machine is adopted to respectively construct the decision hyperplane for each two states, all control parameters are based once, basic control parameters are based once, and finally the two decision hyperplanes are weighted, so that the prediction accuracy of the fault state determination model is improved.

In some embodiments, the set of base control parameters is determined based on correlation coefficients between all control parameters in the set of preset control parameters and a preset correlation coefficient threshold.

The specific determination method is as follows:

because the control parameters of different faults have certain relevance and can be mutually influenced, in order to reduce the dimension of the feature vector formed by the control parameters and reduce the calculated amount, the relevance analysis is required to be carried out on the control parameters in the preset control parameter set, and the basic control parameter set is found out from the preset control parameter set.

Aiming at all control parameters in a preset control parameter set, the control parameters are related to each other, and the related coefficients of the two parameters are calculated based on a related coefficient calculation formula.

The correlation coefficient calculation formula is:

and solving the correlation coefficient among all the control parameters.

And screening out control parameters which are larger than a preset correlation coefficient threshold value according to the correlation coefficient between the control parameters and the preset correlation coefficient threshold value to form a basic control parameter set.

For example, the preset correlation coefficient threshold may be 0.5, and may be set according to different application scenarios, which is not limited herein.

In some embodiments, the preset control parameter set includes a number of missing cigarettes in the hopper branches, a number of damaged cigarettes in the cigarette delivery module, a cigarette uniformity of the cigarette delivery module, a number of aluminum foil papers, a position of the aluminum foil papers, a length of the aluminum foil papers, a wheel feeding rotation speed of the aluminum foil papers, a wheel discharging rotation speed of the aluminum foil papers, a content of the inner frame papers, a content of the trademark papers, a wheel feeding rotation speed of the trademark papers, and a wheel discharging rotation speed of the trademark papers.

Constructing a first decision hyperplane ω x between each two states based on a set of basic control parameters _i +b=0, constructing a second decision hyperplane ω' ·x based on a preset set of control parameters _i +b′＝0。

And respectively constructing a classification hyperplane model for the normal state and the fault state, the normal state and the disturbance state, and the fault state and the disturbance state. The model needs to be trained based on historical operation data, and 6 decision hyperplanes need to be respectively constructed, which are respectively: the method comprises the steps of training a normal fault first decision hyperplane based on a basic control parameter set corresponding to a normal state and a fault state in a training sample, training a normal fault second decision hyperplane based on a preset control parameter set corresponding to the normal state and the fault state in the training sample, training a normal disturbance first decision hyperplane based on the basic control parameter set corresponding to the normal state and the disturbance state in the training sample, training a normal disturbance second decision hyperplane based on the preset control parameter set corresponding to the normal state and the disturbance state in the training sample, training a fault disturbance first decision hyperplane based on the basic control parameter set corresponding to the fault state and the disturbance state in the training sample, and training a fault disturbance second decision hyperplane based on the fault state and the preset control parameter set corresponding to the disturbance state in the training sample.

Taking the normal state and the fault state as examples, the construction of the first decision super for normal fault will be describedA planar process. The method comprises the following steps: y for normal state _i Expressed by =1, the fault state is denoted by y _i = -1, building a first decision hyperplane: omega. X _i +b=0, the positive hyperplane being ω·x _i +b=1, negative hyperplane ω·x _i +b= -1, point-to-plane distance y _i (ω·x _i +b) > 1, the decision function is g (x) =sign (ω·x) _i +b), distance (interval) of positive and negative hyperplanes:

wherein ω= (ω) ₁ ,ω ₂ ,ω ₃ ,...,ω _n ) Is a normal vector of the hyperplane and is also a weight coefficient vector. b is a constant term of hyperplane, x _i Data of a basic control parameter set which is an input training sample set.

The values of omega and b can be solved by training the data of the basic control parameter set of the sample set, and the first decision hyperplane is determined.

Solving the optimal decision hyperplane, namely making the interval as large as possible, and the larger the difference between the two types of data, the more remarkable the classification effect. According to the interval expression, the interval is as large as possible, that is, the normal vector norm is as small as possible. For the hard-spacing problem, solving the decision hyperplane without or without considering outliers can translate into the following:

wherein,

further converting the solution content to obtain

Converting the inequality constraint into an equality constraint to obtain

(s is the number of samples of the training sample set)

Aiming at solving the target to find the extremum, a method for finding the extremum by using a Lagrange function is introduced,

λ _i called Lagrangian multiplier, extremums, the function L is applied to ω, b, λ, respectively _i ,p _i Deriving, the derivative being 0, i.e

The results obtained are as follows:

converting equation 4 intoCombining with formula (3), there is

λ _i [y _i (ω·x _i +b)-1]=0, and λ _i ≥0；

Will Lagrangian multiplier lambda _i And (3) regarding the result as a penalty coefficient against constraint conditions, wherein the result meets KKT conditions, and the optimal decision hyperplane can be solved. The dual problem of the original problem to be solved is:

maximize q(λ _i )＝maximize(minimize(L(ω,b,λ _i ,p _i )))，λ _i ≥0；

when the strong dual condition is satisfied (q (λ) _i ) ^* ＝f(ω) ^* ) When the convex optimization problem is solved under the constraint of the current affine function, the original problem and the dual problem simultaneously obtain the optimal solution。

The dual problem is simplified to obtain

Solving lambda _i After the value of (2)Obtaining omega; then according to lambda _i [y _i (ω·x _i +b)-1]=0, and support vectors brought into the positive and negative classes solve for the value of b. By introducing the dual problem, the calculation process is greatly simplified.

For the situation that abnormal values occur, a loss function and a soft interval are defined, and the solving problem of the optimal decision hyperplane in the situation can be obtained through the same processing method, wherein the solving target is as follows:

Solving the target:

wherein ε _i ＝max(0,1-y _i (ω·x _i +b)) is a loss function; c is a super parameter, the size of which controls the tolerance to the loss function, and the larger the C, the lower the tolerance, the smaller the loss function.

If the problem is nonlinear, firstly, the data of the original parameter set is mapped to a new space by using a kernel transformation, and then, the subsequent classification model training is carried out in the new space by using a linear classification method. Wherein x is the data of the original parameter set, and the new feature vector obtained by mapping after nuclear transformation is recorded asThe decision hyperplane of the feature space can be expressed as +.>

Normal fault second decision hyperplane ω '·xi' +b '=0, where xi' is the data of the preset control parameter set of the training sample set. The calculation of ω 'and b' of the second decision hyperplane is not described in detail here.

The first decision hyperplane of the normal fault and the second decision hyperplane of the normal fault are weighted to obtain the first decision hyperplane, and the specific weighting process is as follows:

weighting the omega and omega ' values and weighting the b and b ' values to obtain omega ' =alpha corresponding to the basic control parameter set ₁ ω+α ₂ ω′，b″＝α ₁ b+α ₂ b', wherein alpha ₁ +α ₂ ＝1，α ₁ 、α ₂ The weights of the two models are set manually. For omega' =alpha corresponding to other control parameters except the basic control parameter set in the preset control parameter set ₁ ω+α ₂ Omega', where alpha ₁ ＝0,α ₂ =1, ω=0. The above operation is equivalent to increasing the influence of the basic fault numerical characteristics so as to realize higher classification accuracy.

The first decision hyperplane finally obtained is:

ω″·x _i +b″＝0，

wherein,

then

Wherein alpha is ₁ +α ₂ =1, which are weights of two models set by human, respectively, to maximize classification accuracy of the models.

Note that ω and ω' are not one numerical value but one numerical sequence.

The construction process of the second decision hyperplane and the third decision hyperplane is not described here in detail, and is the same as the construction process of the first decision hyperplane, except that the data used in the construction process is different.

The final fault state determination model includes a first decision hyperplane, a second decision hyperplane, and a third decision hyperplane. Each decision hyperplane outputs an operation state corresponding to each data according to the input data.

After the training of the fault state determination model is completed, the trained fault state determination model also needs to be tested by using the test data set, so that the accuracy of the model is determined.

After the fault state determination model is trained, the model can be used for detecting the operation data.

After the fault state determining model is trained, the model can be used for detecting the running data for a period of time.

Because the fault state determination model comprises 3 decision hyperplanes, the input operation data of each period need to be respectively input into the 3 decision hyperplanes, the 3 decision hyperplanes output an operation state, at the moment, all the output operation states need to be counted, the operation state with the largest output times is found, and the operation state with the largest output times is determined as the final operation state of the packaging machine. If the times of the three operation states are the same, the fault state is preferentially used as the final output result, and the packaging machine is prevented from generating unnecessary loss. If the number of times of the two operation states is the same, the result is output with the serious state of the two states as the priority.

Such as acquisition day 14: and (3) collecting all operation data sets between 00 and 14:30 by the packaging machine every 5 minutes, wherein 6 groups of operation data are generated, the 6 groups of operation data are input into a fault state determination model, 18 operation states are output by the 3 decision hyperplane, and the operation state with the largest output times is finally taken as a final output result by the fault state determination model to serve as the current operation state of the packaging machine. If the times of the three operation states are the same, the fault state is preferentially used as the final output result, and the packaging machine is prevented from generating unnecessary loss. If the number of times of the two operation states is the same, the result is output with the serious state of the two states as the priority.

In some embodiments, if the detected result is a normal state, a feedback signal is sent for normal operation of the packaging machine. And if the detected result is in a fault state, sending a feedback signal of the fault of the cigarette packaging machine. When detecting that the cigarette packing machine is in a fault state, the fault type needs to be further analyzed, and the specific analysis process is as follows:

step S1211, a target operation data set of the preset control parameter set is subjected to normalization processing.

And S1212, comparing the standardized operation data set with a standard data set, and determining a significant control parameter set deviating from a normal working state.

The salient control parameter set comprises at least one control parameter.

All data in the standard data set are also data obtained after normalization processing.

The standard value for each control parameter may be determined based on an average value in all normal run data sets in the training sample. Or the standard value of each control parameter is set by a professional according to working experience.

After determining the actual value and the standard value of each control parameter, the deviation degree of the control parameter can be determined. Degree of deviation theta _i The calculation mode of (a) is as follows:

Since the collected operation data set may include data of a plurality of time periods, that is, a plurality of groups of control parameters, in order to further determine the accuracy of calculation, the deviation degree of each control parameter in each group of control parameters may be first obtained, and then the deviation degree of each control parameter in all the obtained groups of control parameters may be averaged. Of course, it may be determined by other calculation methods according to the actual application scenario, which is not limited herein.

The above description is given by way of example: for the acquired 6 groups of data, the deviation degree of each group of control parameters can be firstly calculated, and then the 6 deviation degrees of each control parameter in the 6 groups of data are averaged, namely the final ith control parameter deviation degree is

And finally, based on the average deviation degree of all the control parameters in a certain period, sorting the average deviation degree of all the control parameters, and selecting a plurality of control parameters with larger average deviation degree in the prior sorting as a significant control parameter set.

For example, the control parameters corresponding to the first 3 largest average deviations may be selected.

Step S1213, searching is performed based on the significant control parameter set and the pre-constructed fault tree, and the fault type corresponding to the fault state is determined.

The fault tree may be constructed from the operational data set at the time of the fault condition in the historical parameters and the reasons for the fault and the associated fault solutions recorded by the staff. The operational data in the fault condition may be stored in a hierarchical manner.

The fault tree includes fault models, fault subsystems, fault locations, fault types, fault typical phenomena, cigarette brands, solutions, and the like.

For example, taking a ZB45 type hard box packaging machine as an example, wherein the fault machine type is the ZB45 type hard box packaging machine; the fault subsystem is an out-of-strip transparent paper packaging machine, a hard strip packaging machine, an out-of-box transparent paper packaging machine and a hard box packaging machine; taking a hard box packaging machine as an example, the fault parts corresponding to possible faults under the subsystem comprise four-wheel part faults, five-wheel part faults and six-wheel part faults; among the types of faults that may occur in the five-wheel region are: fly inner frame paper, frequent alarm, label paper blockage, label paper suction failure and the like; the reasons or characteristics of the blockage faults of the trademark paper are that the negative pressure is too small, the trademark paper deforms, the position of the paper warehouse is inaccurate, and the like, wherein the solution corresponding to the too small negative pressure is to increase the negative pressure to a proper value.

The set of significant control parameters is retrieved from the fault tree, and then the fault tree is retrieved from each control parameter in the set of significant control parameters, matching the most likely fault type and solution.

Then, a weight table is constructed according to the matched fault types and the significant control parameter sets. Finally, determining the most probable fault type according to the constructed weight table, and then giving a related solution by combining the opinion of the staff. The following is a specific example:

assuming that the significant control parameter set has four control parameters a, b, c and D, assuming that the most likely fault types to which the 4 control parameters match are A, B, C and D, respectively, a weight table is constructed from the 4 control parameters and the 4 fault types.

The weight table may be constructed according to historical operation data, for example, 100 faults occur in total in a set period, 50 faults occur in the type a, 30 faults occur in the type B, and 10 faults occur in the type C and the type D. Then, the times of faults of each control parameter under each fault type are counted respectively. For example, the number of failures of the a control parameter in the a failure type is 10 times, the number of failures of the b control parameter in the a failure type is 15 times, the number of failures of the c control parameter in the a failure type is 25 times, and the number of failures of the d control parameter in the a failure type is 0 times. The number of times of failure of the a control parameter in the B failure type is 5 times, the number of times of failure of the B control parameter in the B failure type is 15 times, the number of times of failure of the c control parameter in the B failure type is 0 times, and the number of times of failure of the d control parameter in the B failure type is 10 times. The number of times of failure of the a control parameter in the C failure type is 0 times, the number of times of failure of the b control parameter in the C failure type is 5 times, the number of times of failure of the C control parameter in the C failure type is 5 times, and the number of times of failure of the d control parameter in the C failure type is 0 times. The number of times of failure of the a control parameter in the D failure type is 3 times, the number of times of failure of the b control parameter in the D failure type is 2 times, the number of times of failure of the c control parameter in the D failure type is 4 times, and the number of times of failure of the D control parameter in the D failure type is 1 time. The weight table constructed is shown in table 1 below:

Table 1 weight table

	a	b	c	d
					A	0.1	0.15	0.25	0
B	0.05	0.15	0	0.1
					C	0	0.05	0.05	0
D	0.03	0.02	0.04	0.01

And selecting the first few fault types with higher weight as typical fault types according to the weight table, and then combining the opinion of staff to give related solutions.

Finally, the packaging machine is maintained according to the finally determined typical fault type and solution.

In some embodiments, when it is determined that the target cigarette packer is in a disturbance state, a feedback signal of the disturbance of the cigarette packer is sent. When the packing machine is detected to be in a disturbance state, which indicates that the packing machine is slightly abnormal, the running trend of the packing machine needs to be further predicted so as to perform corresponding processing. The specific treatment process is as follows:

and based on the current operation data and a time sequence corresponding to the operation data set, inputting the current operation data and the time sequence corresponding to the operation data set into a trained time sequence prediction model, and predicting the operation data set of the target cigarette packaging machine in a preset period.

In this embodiment, the time series prediction model is an ARIMA model, and the model is constructed as follows:

step S1221, forming a time series.

Historical data are collected, a time sequence is formed for the operation data of each control parameter respectively, the time sequence of each control parameter is drawn into a time sequence diagram, curve trend is observed, and whether the time sequence is a stable sequence is judged. The mean value and the variance of the stable sequence are judged to be constants, and the corresponding time sequence diagram shows that the sequence randomly fluctuates around a constant value all the time, and the fluctuation range is bounded and has no obvious trend and periodic characteristics. If higher accuracy is desired, verification can be performed using the root of unity test method (ADF test). Judging whether a time sequence has a unit root or not, if so, proving that the sequence is unstable, and enabling pseudo regression to exist in regression analysis; operations may be accomplished by SPSS software.

Step S1222, converting the non-stationary sequence into a stationary sequence by a difference method.

If the detected time sequence is a non-stationary sequence, converting the time sequence into a stationary sequence by a difference method, wherein the specific formula is as follows: Δy _x The number of differences d can be determined by a numerical graph after the difference, as in the method of determining stationarity above, until a stationary sequence of a larger interval occurs.

Step S1223, determining the values of p and q in ARIMA (p, d, q).

The d value has been determined in step S1223, and the values of p, q need to be determined by introducing the autocorrelation coefficients ACF and the partial autocorrelation coefficients PACF.

ACF reflects the correlation between the values of the same sequence at different time sequences, PACF calculates the correlation between two strict variables, and the correlation degree between the two variables obtained after the interference of the intermediate variables is eliminated. For time series X _t ：X ₁ 、X ₂ 、X _3...... X _t-1 、X _t 。

ACF：X _t And X is _t-k The "correlation coefficient" of (1) is X _t The autocorrelation coefficients of interval k;

PACF: post-interference X with intermediate k-1 random variables _t-k For X _t Correlation metrics of influence.

After obtaining ACF and PACF images, the observation patterns determine the values of p and q: the auto. ARIMA function in R language may be used to perform automatic order determination, and then, in combination with its own judgment, several different p, q are selected, which are generally smaller than 5, such as ARIMA (2, 1), ARIMA (2, 2), etc., and then, the AIC criterion is used to determine the optimal model.

AIC red pool information criterion, parameter and result precision trade-off; aic=2k-2 ln (L), k being the number of model parameters and L being the likelihood function.

After the optimal model is determined, the model needs to be checked. The test procedure was as follows: first, the significance of the parameter estimation is checked using t-test. Then, checking the randomness of the residual sequence, and checking the residual randomness of the optimal ARIMA (p, d, q) model by an autocorrelation function method to make an autocorrelation function diagram of the residual. If the residual independence is high through inspection, the autocorrelation does not exist, and the model is available.

After the time sequence prediction model is constructed and detected, the model can be used for short-term prediction, and the operation data within 5-10 minutes in the future can be obtained.

After the predicted operation data is obtained, the predicted operation data is input into a fault state determination model to predict the operation state of the target cigarette packing machine. And determining whether the target cigarette packaging machine maintains the interference state or can develop into the fault state according to the prediction result, thereby realizing daily operation and maintenance and pre-maintenance of the packaging machine. If the state is still disturbance, the running state of the packaging machine in the period of time needs to be paid attention to all the time. If the fault state is predicted, normal maintenance is carried out on the packaging machine after the machine is stopped.

The method for determining the running state of the cigarette packing machine comprises the steps of firstly, acquiring a running data set of a target cigarette packing machine in a current period, and then, inputting the running data set into a trained fault state determining model to determine the running state of the target cigarette packing machine. The fault state determination model constructed by the invention is obtained by carrying out weighting treatment on a first decision hyperplane trained based on a preset control parameter set and a second decision hyperplane trained based on a basic control parameter set, and the model can more accurately predict the running state of the cigarette packaging machine. The method for determining the running state of the cigarette packing machine can determine the current running state of the cigarette packing machine in time and accurately according to the running data set of the current time period, can greatly reduce the loss caused by untimely fault clearing of the cigarette packing machine, and is easy to realize the pre-maintenance and normal operation and maintenance of the cigarette packing machine.

It should be understood that the sequence number of each step in the foregoing embodiment does not mean that the execution sequence of each process should be determined by the function and the internal logic, and should not limit the implementation process of the embodiment of the present invention.

Based on the method for determining the running state of the cigarette packaging machine provided by the embodiment, correspondingly, the invention further provides a specific implementation mode of the device for determining the running state of the cigarette packaging machine, which is applied to the method for determining the running state of the cigarette packaging machine. Please refer to the following examples.

As shown in fig. 2, there is provided a determining apparatus 200 of an operation state of a cigarette packing machine, the apparatus comprising:

an obtaining module 210, configured to obtain a target operation data set of the target cigarette packaging machine in a current period, where the target operation data set is a data set obtained by detecting a preset control parameter set in real time when the target cigarette packaging machine is operated, and the preset control parameter set includes a plurality of control parameters;

a determining module 220, configured to input the target operation data set into a trained fault state determining model, so as to determine an operation state of the target cigarette packaging machine; the fault state determining model is obtained based on a first decision hyperplane and a second decision hyperplane, the first decision hyperplane is constructed based on a first training sample set formed by basic control parameter sets of training samples, the second decision hyperplane is constructed based on a second training sample set formed by preset control parameter sets of training samples, the basic control parameter sets are extracted from the preset control parameter sets according to preset extraction rules, the training samples comprise an operation data set in normal operation, an operation data set in disturbance state and an operation data set in fault state, each data set is provided with a corresponding data set state label, and the operation states comprise a normal state, a disturbance state and a fault state.

In one possible implementation, the determining module 220 is configured to send a feedback signal of the failure of the cigarette packaging machine when it is determined that the target cigarette packaging machine is in the failure state;

a determining module 220, configured to determine a deviation of the target control parameter based on the actual value of the target control parameter and a standard value of the target control parameter, where the target control parameter is any one control parameter;

degree of deviation theta _i The calculation mode of (a) is as follows:

In one possible implementation, the determining module 220 is configured to send a feedback signal of disturbance of the cigarette packaging machine when it is determined that the target cigarette packaging machine is in a disturbance state;

inputting the target operation data set and the time sequence corresponding to the operation data set into a trained time sequence prediction model, and predicting the operation data set of the target cigarette packaging machine in a preset period;

the preset control parameter set comprises the number of cigarettes missing of a hopper branch, the number of damaged cigarettes in a cigarette delivering mould box, the uniformity of the cigarettes in the cigarette delivering mould box, the number of aluminum foil papers, the position of the aluminum foil papers, the length of the aluminum foil papers, the rotating speed of an aluminum foil cigarette packet feeding wheel, the rotating speed of an aluminum foil cigarette packet discharging wheel, the content of inner frame papers, the content of trademark papers, the rotating speed of a trademark paper feeding wheel and the rotating speed of a trademark paper discharging wheel;

Fig. 3 is a schematic diagram of an electronic device according to an embodiment of the present invention. As shown in fig. 3, the electronic apparatus 3 of this embodiment includes: a processor 30, a memory 31 and a computer program 32 stored in said memory 31 and executable on said processor 30. The processor 30, when executing the computer program 32, implements the steps of the above-described embodiments of the method for determining the operating state of each cigarette packer, such as steps 110 to 120 shown in fig. 1. Alternatively, the processor 30 may perform the functions of the modules of the apparatus embodiments described above, such as the functions of the modules 210-220 of fig. 2, when executing the computer program 32.

Illustratively, the computer program 32 may be partitioned into one or more modules that are stored in the memory 31 and executed by the processor 30 to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing the specified functions for describing the execution of the computer program 32 in the electronic device 3. For example, the computer program 32 may be partitioned into modules 210 through 220 shown in FIG. 2.

The electronic device 3 may include, but is not limited to, a processor 30, a memory 31. It will be appreciated by those skilled in the art that fig. 3 is merely an example of the electronic device 3 and does not constitute a limitation of the electronic device 3, and may include more or fewer components than shown, or may combine certain components, or different components, e.g., the electronic device may further include an input-output device, a network access device, a bus, etc.

The processor 30 may be a central processing unit (Central Processing Unit, CPU), other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), field-programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

The memory 31 may be an internal storage unit of the electronic device 3, such as a hard disk or a memory of the electronic device 3. The memory 31 may be an external storage device of the electronic device 3, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card) or the like, which are provided on the electronic device 3. Further, the memory 31 may also include both an internal storage unit and an external storage device of the electronic device 3. The memory 31 is used for storing the computer program and other programs and data required by the electronic device. The memory 31 may also be used for temporarily storing data that has been output or is to be output.

It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-described division of the functional units and modules is illustrated, and in practical application, the above-described functional distribution may be performed by different functional units and modules according to needs, i.e. the internal structure of the apparatus is divided into different functional units or modules to perform all or part of the above-described functions. The functional units and modules in the embodiment may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit, where the integrated units may be implemented in a form of hardware or a form of a software functional unit. In addition, specific names of the functional units and modules are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working process of the units and modules in the above system may refer to the corresponding process in the foregoing method embodiment, which is not described herein again.

In the foregoing embodiments, the descriptions of the embodiments are emphasized, and in part, not described or illustrated in any particular embodiment, reference is made to the related descriptions of other embodiments.

Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

In the embodiments provided in the present invention, it should be understood that the disclosed apparatus/electronic device and method may be implemented in other manners. For example, the apparatus/electronic device embodiments described above are merely illustrative, e.g., the division of the modules or units is merely a logical function division, and there may be additional divisions in actual implementation, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection via interfaces, devices or units, which may be in electrical, mechanical or other forms.

The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.

The integrated modules/units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the present invention may implement all or part of the above-described methods, or may be implemented by a computer program for instructing related hardware, where the computer program may be stored in a computer readable storage medium, and the computer program may be executed by a processor to implement the steps of the above-described method embodiments for determining the operation state of each cigarette packer. Wherein the computer program comprises computer program code which may be in source code form, object code form, executable file or some intermediate form etc. The computer readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), an electrical carrier signal, a telecommunications signal, a software distribution medium, and so forth.

The above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention, and are intended to be included in the scope of the present invention.

Claims

1. A method for determining the operating condition of a cigarette packer, comprising:

inputting the target operation data set into a trained fault state determination model to determine the operation state of the target cigarette packaging machine; the fault state determining model is obtained based on a first decision hyperplane and a second decision hyperplane through weighting processing, the first decision hyperplane is constructed based on a first training sample set formed by basic control parameter sets of training samples, the second decision hyperplane is constructed based on a second training sample set formed by preset control parameter sets of the training samples, the basic control parameter sets are extracted from the preset control parameter sets according to preset extraction rules, the training samples comprise an operation data set in normal operation, an operation data set in disturbance state and an operation data set in fault state, each data set is provided with a corresponding data set state label, and the operation states comprise a normal state, a disturbance state and a fault state.

2. The method of determining the operating condition of a cigarette packer of claim 1, wherein the method further comprises:

when the target cigarette packing machine is determined to be in a fault state, a feedback signal of the packing machine fault is sent out;

comparing the target operation data set after the standardization processing with a standard data set, and determining a significant control parameter set deviating from a normal working state, wherein the significant control parameter set at least comprises one control parameter;

3. The method for determining the operating state of a cigarette packing machine according to claim 2, wherein the standard data set includes a standard value of each control parameter;

comparing the target operation data set after the standardization processing with a standard data set, and determining a significant control parameter set deviating from a normal working state, wherein the method comprises the following steps:

determining the deviation degree of a target control parameter based on the actual value of the target control parameter and the standard value of the target control parameter, wherein the target control parameter is any one control parameter;

the degree of deviation theta _i The calculation mode of (a) is as follows:

4. A method of determining the operating condition of a cigarette packer as in claim 3, wherein the standard value of the target control parameter is determined based on an average value of the target control parameter in all operating data sets during normal operation in the training sample.

5. The method of determining the operating condition of a cigarette packer of claim 1, wherein the method further comprises:

when the target cigarette packing machine is determined to be in a disturbance state, sending a feedback signal of packing machine disturbance;

and inputting the operation data set in the preset period into the fault state determining model to determine the operation state trend of the target cigarette packaging machine.

6. The method for determining the operating state of a cigarette packer according to any one of claims 1 to 5, wherein the basic control parameter set is determined based on a correlation coefficient between all control parameters in the preset control parameter set and a preset correlation coefficient threshold;

the preset control parameter set comprises the number of cigarettes lacking of a hopper branch, the number of damaged cigarettes in a cigarette delivering mould box, the number of cigarettes continuously removed in the cigarette delivering mould box, the number of aluminum foil papers, the position of the aluminum foil papers, the length of the aluminum foil papers, the rotation speed of an aluminum foil cigarette packet feeding wheel, the rotation speed of an aluminum foil cigarette packet discharging wheel, the content of inner frame papers, the content of trademark papers, the rotation speed of a trademark paper feeding wheel and the rotation speed of a trademark paper discharging wheel;

the basic control parameter set comprises the number of cigarettes lacking of a hopper branch, the number of damaged cigarettes in a cigarette delivering mould box, the number of continuous cigarettes removing in the cigarette delivering mould box, the number of aluminum foil papers, the position of the aluminum foil papers, the rotating speed of an aluminum foil cigarette packet feeding wheel, the content of inner frame papers, the content of label papers and the rotating speed of a label paper feeding wheel.

7. The method for determining the operating state of a cigarette packer according to any one of claims 1 to 5, wherein the fault state determination model includes 3 decision hyperplanes, a first decision hyperplane being weighted by a normal fault first decision hyperplane constructed based on a normal state in the training sample and a corresponding basic control parameter set in the fault state and a normal fault second decision hyperplane constructed based on a normal state in the training sample and a corresponding preset control parameter set in the fault state; the second decision hyperplane is obtained by weighting a normal disturbance first decision hyperplane constructed based on a corresponding basic control parameter set in a normal state and a disturbance state in the training sample and a normal disturbance second decision hyperplane constructed based on a corresponding preset control parameter set in the normal state and the disturbance state in the training sample; the third decision hyperplane is obtained by weighting a first decision hyperplane of fault disturbance constructed based on a corresponding basic control parameter set in a fault state and a disturbance state in the training sample and a second decision hyperplane of fault disturbance constructed based on a corresponding preset control parameter set in the fault state and the disturbance state in the training sample;

The inputting the target operation data set into a trained fault state determination model to determine an operation state of the target cigarette packaging machine comprises the following steps:

and determining the most classification result in all the classification results as the running state of the target cigarette packaging machine.

8. A device for determining the operating condition of a cigarette packer, comprising:

the system comprises an acquisition module, a control module and a control module, wherein the acquisition module is used for acquiring a target operation data set of a target cigarette packaging machine in a current period, wherein the target operation data set is obtained by detecting a preset control parameter set in real time when the target cigarette packaging machine operates, and the preset control parameter set comprises a plurality of control parameters;

the determining module is used for inputting the target operation data set into a trained fault state determining model so as to determine the operation state of the target cigarette packaging machine; the fault state determining model is obtained based on a first decision hyperplane and a second decision hyperplane through weighting processing, the first decision hyperplane is constructed based on a first training sample set formed by basic control parameter sets of training samples, the second decision hyperplane is constructed based on a second training sample set formed by preset control parameter sets of the training samples, the basic control parameter sets are extracted from the preset control parameter sets according to preset extraction rules, the training samples comprise an operation data set in normal operation, an operation data set in disturbance state and an operation data set in fault state, each data set is provided with a corresponding data set state label, and the operation states comprise a normal state, a disturbance state and a fault state.

9. An electronic device comprising a memory for storing a computer program and a processor for invoking and running the computer program stored in the memory to perform the method of any of claims 1 to 7.

10. A computer readable storage medium storing a computer program, characterized in that the computer program when executed by a processor implements the steps of the method according to any one of claims 1 to 7.