Force feedback control method and system based on digital cash register buttons

Abstract

The present invention discloses a force feedback control method and system based on digital cash register keys, and relates to the field of electronic equipment, wherein the method comprises: collecting digital information and bit information currently input by a user through a digital cash register key; obtaining unit abnormality parameters through an error analysis module; judging whether the user has previously input previous digital information according to the bit information; training a continuous identifier to identify at least one previous digital information, at least one previous bit information, digital information and bit information, and obtaining continuous abnormality parameters; obtaining comprehensive abnormality parameters through weighted calculation; classifying the comprehensive abnormality parameters into input abnormality levels, and obtaining input abnormality levels and force feedback levels; and performing key force feedback according to the force feedback levels through a force feedback control module. The method solves the technical problems of low key abnormality recognition accuracy of digital cash registers and poor key force feedback effect of digital cash registers in the prior art.

Description

Technical Field

[0001] The present invention relates to the field of electronic equipment, and in particular to a force feedback control method and system based on keys of a digital cash register. Background Art

[0002] With the continuous intelligent development of society, the popularity of digital cash registers continues to expand. Digital cash registers have the advantages of saving labor, saving time, and low cost. Key force feedback has an important impact on the normal use of digital cash registers. In the prior art, there are technical problems such as low accuracy in abnormal key recognition of digital cash registers and poor key force feedback effect of digital cash registers. Summary of the invention

[0003] The present application provides a force feedback control method and system based on digital cash register keys. The present application solves the technical problems of low accuracy in identifying abnormal keys of digital cash registers and poor force feedback effect of keys of digital cash registers in the prior art. The present application achieves the technical effects of improving the accuracy in identifying abnormal keys of digital cash registers, improving the force feedback effect of keys of digital cash registers, and improving the intelligence of force feedback of keys of digital cash registers.

[0004] In view of the above problems, the present application provides a force feedback control method and system based on digital cash register buttons.

[0005] In a first aspect, the present application provides a force feedback control method based on digital cash register buttons, wherein the method is applied to a force feedback control system based on digital cash register buttons, the system is communicatively connected to a force feedback control device based on digital cash register buttons, the device comprises an input acquisition module, an error analysis module, a force feedback control module and a plurality of feedback components, the plurality of feedback components are arranged on a plurality of button components in the digital cash register, the method comprises: collecting, by the input acquisition module, digital information and bit information currently input by a user through the digital cash register buttons as current input; training a unit identifier by the error analysis module based on historical input data of the digital cash register or the user, identifying the digital information and bit information, and obtaining unit abnormality parameters; According to the position information, determine whether the user has previously input previous digital information. If so, extract and obtain at least one previous digital information and at least one previous position information. If not, calculate the comprehensive abnormality parameter according to the unit abnormality parameter; based on the historical input data of the digital cash register or the user, train a continuous identifier to identify at least one previous digital information, at least one previous position information, digital information and position information to obtain continuous abnormality parameters; according to the continuous abnormality parameters and the unit abnormality parameters, perform weighted calculation to obtain the comprehensive abnormality parameters; classify the comprehensive abnormality parameters into input abnormality levels to obtain input abnormality levels and force feedback levels; through the force feedback control module, perform key force feedback according to the force feedback level through the feedback component on the key component corresponding to the digital information.

[0006] In a second aspect, the present application further provides a force feedback control system based on digital cash register buttons, wherein the system is communicatively connected to a force feedback control device based on digital cash register buttons, the device comprising an input acquisition module, an error analysis module, a force feedback control module and a plurality of feedback components, the plurality of feedback components being arranged on a plurality of button components in the digital cash register, the system comprising: a current input acquisition module, the current input acquisition module being used to collect the digital information and bit information currently input by the user through the digital cash register buttons as the current input through the input acquisition module; a unit abnormality recognition module, the unit abnormality recognition module being used to train a unit identifier through the error analysis module based on the historical input data of the digital cash register or the user, to identify the digital information and bit information, and to obtain unit abnormality parameters; a previous position judgment module, the previous position judgment module being used to judge whether the user has previously input a previous digital information according to the bit information. information, if yes, extract and obtain at least one previous digital information and at least one previous position information, if not, calculate the comprehensive abnormality parameter according to the unit abnormality parameter; a continuous abnormality recognition module, the continuous abnormality recognition module is used to train a continuous identifier based on the historical input data of the digital cash register or the user, identify at least one previous digital information, at least one previous position information, digital information and position information, and obtain continuous abnormality parameters; a comprehensive abnormality calculation module, the comprehensive abnormality calculation module is used to obtain comprehensive abnormality parameters by weighted calculation according to the continuous abnormality parameters and unit abnormality parameters; an abnormality level classification module, the abnormality level classification module is used to classify the input abnormality level of the comprehensive abnormality parameters, and obtain the input abnormality level and force feedback level; a force feedback module, the force feedback module is used to perform key force feedback according to the force feedback level through the feedback component on the key component corresponding to the digital information through the force feedback control module.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] The digital information and bit information currently input by the user through the digital cash register keys are collected through the input acquisition module; the digital information and bit information are identified through the error analysis module to obtain the unit abnormality parameter; it is determined whether the user has previously input the previous digital information, if so, at least one previous digital information and at least one previous bit information are extracted, if not, the comprehensive abnormality parameter is calculated according to the unit abnormality parameter; the continuous identifier is trained to identify at least one previous digital information, at least one previous bit information, digital information and bit information to obtain the continuous abnormality parameter; the comprehensive abnormality parameter is obtained by weighted calculation according to the continuous abnormality parameter and the unit abnormality parameter; the comprehensive abnormality parameter is classified into the input abnormality level to obtain the input abnormality level and the force feedback level; the feedback component on the key component corresponding to the digital information is fed back according to the force feedback level. The technical effect of improving the key abnormality recognition accuracy of the digital cash register, improving the key force feedback effect of the digital cash register, and improving the key force feedback intelligence of the digital cash register is achieved.

[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings of the embodiments of the present invention are briefly introduced below. Obviously, the drawings in the following description only relate to some embodiments of the present invention, but are not intended to limit the present invention.

[0011] Figure 1 A schematic flow chart of a force feedback control method based on digital cash register buttons in this application;

[0012] Figure 2 A schematic diagram of a flow chart of input abnormality level classification of comprehensive abnormal parameters in a force feedback control method based on digital cash register buttons in the present application;

[0013] Figure 3 This is a structural schematic diagram of a force feedback control system based on digital cash register buttons in this application. DETAILED DESCRIPTION

[0014] The present application provides a force feedback control method and system based on digital cash register keys, which solves the technical problems of low accuracy in identifying abnormal keys of digital cash registers and poor force feedback effect of keys of digital cash registers in the prior art, and achieves the technical effects of improving the accuracy in identifying abnormal keys of digital cash registers, improving the force feedback effect of keys of digital cash registers, and improving the intelligence of key force feedback of digital cash registers.

[0015] Embodiment 1

[0016] Please refer to the attached Figure 1 The present application provides a force feedback control method based on digital cash register buttons, wherein the method is applied to a force feedback control system based on digital cash register buttons, the system is communicatively connected with a force feedback control device based on digital cash register buttons, the device comprises an input acquisition module, an error analysis module, a force feedback control module and a plurality of feedback components, the plurality of feedback components are arranged on a plurality of button components in the digital cash register, the method specifically comprises the following steps:

[0017] The input collection module collects the digital information and bit information currently input by the user through the keys of the digital cash register as the current input;

[0018] Connect the input acquisition module, collect the digital information and bit information currently input by the user through the digital cash register keys through the input acquisition module, and set the digital information and bit information currently input by the user through the digital cash register keys as the current input. The input acquisition module has the function of sensing and reading the input information of the digital cash register keys. The digital information includes the specific number currently input by the user through the digital cash register keys. For example, the digital information includes 0, 1, 2, 3, 4, etc. The bit information includes the digit corresponding to the specific number currently input by the user through the digital cash register keys, and the number corresponding to the specific number. For example, the digit includes the ones, tens, hundreds, etc. The number includes whether the specific number currently input is the first digit / the specific number currently input is the second digit, etc.

[0019] Through the error analysis module, based on the historical input data of the digital cash register or the user, a unit identifier is trained to identify the digital information and the bit information to obtain unit abnormality parameters;

[0020] Based on the login record in the digital cash register, obtaining the user record;

[0021] determining whether there is more than one user in the user record, and if not, training the unit identifier based on historical input data in the digital cash register;

[0022] If yes, then obtain the login time records of multiple users, and determine whether the login time difference between the maximum login time and the minimum login time is greater than the login time difference threshold according to the login time records;

[0023] If so, the unit identifier is trained using the historical input data of the digital cash register; if not, the unit identifier is trained based on the historical input data of the user.

[0024] Retrieve the login records in the digital cash register to obtain the user records. The user records include multiple historical login records of the digital cash register. Each historical login record includes the historical login time of the digital cash register, the historical login user name, the historical login duration, etc.

[0025] Further, it is determined whether there is more than one user in the user record. If the historical login user names in the user record are the same, then there is no more than one user in the user record, and the unit identifier is trained based on the historical input data of the digital cash register.

[0026] On the contrary, if the historical login user names in the user record are not the same, then there is more than one user in the user record, and multiple user login time records are extracted according to the historical login user names. Each user login time record includes multiple historical login times corresponding to the same historical login user name in the user record. Then, the multiple historical login times in each user login time record are added and calculated to obtain the login time records of multiple users. The login time record of each user includes the sum of the multiple historical login times in each user login time record. Furthermore, the maximum value in the login time records of multiple users is recorded as the maximum login time, the minimum value in the login time records of multiple users is recorded as the minimum login time, the difference between the maximum login time and the minimum login time is recorded as the login time difference, and it is determined whether the login time difference is greater than the login time difference threshold value pre-set and determined by the force feedback control system based on the digital cash register buttons.

[0027] If the login time difference is greater than the login time difference threshold, the unit identifier is trained using the historical input data of the digital cash register. Conversely, if the login time difference is less than / equal to the login time difference threshold, the unit identifier is trained based on the user's historical input data.

[0028] According to the historical input data of the digital cash register or the user, a plurality of sample input digital information sets of a plurality of sample bit information within a plurality of preset historical time periods are obtained;

[0029] According to the multiple sample input digital information sets, the probability of occurrence of the multiple sample digital information in the multiple sample bit information is calculated to obtain the multiple sample occurrence probability sets;

[0030] According to the occurrence probability sets of multiple samples, multiple sample unit abnormal parameter sets are calculated and obtained;

[0031] According to multiple preset historical time periods, the historical input data of the digital cash register or the historical input data of the user are retrieved to obtain multiple sample input digital information sets corresponding to multiple sample position information. The historical input data of the digital cash register / the historical input data of the user both include multiple sample input digital information sets corresponding to multiple sample position information within multiple preset historical time periods. Each preset historical time period includes historical time range information pre-set and determined by the force feedback control system based on the digital cash register buttons. Multiple sample position information includes multiple historical digits. Multiple historical digits include ones, tens, hundreds, etc. Each sample input digital information set includes multiple sample digital information corresponding to each sample position information within multiple preset historical time periods. Multiple sample digital information includes 1, 2, 3, 4, etc.

[0032] Further, according to the multiple sample input digital information sets, the probability of occurrence of multiple sample digital information in each sample bit information is calculated respectively to obtain multiple sample occurrence probability sets. Each sample occurrence probability set includes the probability of occurrence of multiple sample digital information in each sample bit information. Exemplarily, when obtaining multiple sample occurrence probability sets, the ratio of the total number of occurrences of each sample digital information to the total number of multiple sample digital information in each sample input digital information set is set as the probability of occurrence of each sample digital information, thereby obtaining multiple sample occurrence probability sets corresponding to the multiple sample input digital information sets.

[0033] Further, the probability of occurrence of each sample digital information in each sample occurrence probability set is calculated by difference with 1, and multiple sample unit abnormal parameter sets are obtained. Each sample unit abnormal parameter set includes multiple sample unit abnormal parameters corresponding to the probability of occurrence of multiple sample digital information in each sample occurrence probability set. Each sample unit abnormal parameter includes the difference between 1 and the probability of occurrence of each sample digital information.

[0034] Using a plurality of sample digital information and a plurality of sample unit abnormal parameter sets respectively, the unit identifier is trained, wherein the unit identifier includes a plurality of unit identification channels corresponding to the plurality of sample bit information;

[0035] Based on machine learning, multiple unit recognition channels corresponding to multiple sample bit information are constructed respectively;

[0036] Taking digital information as input and unit abnormality parameters as output, respectively using the plurality of sample digital information and the plurality of sample unit abnormality parameter sets to train the plurality of unit recognition channels until convergence;

[0037] The multiple unit identification channels are integrated to obtain the unit identifier.

[0038] The unit identification channel corresponding to the bit information in the unit identifier is used to identify the digital information to obtain the unit abnormality parameter.

[0039] Machine learning is a science and technology about data learning. Machine learning builds a model based on sample data to predict future behavioral results and trends. Based on machine learning, each unit recognition channel corresponding to each sample bit information is constructed separately, that is, the digital information is used as input information, and the unit abnormality parameter is used as output information. The sample input digital information set and the sample unit abnormality parameter set corresponding to each sample bit information are continuously self-trained and learned until convergence state, so that each unit recognition channel corresponding to each sample bit information can be obtained, and multiple unit recognition channels are added to the unit recognizer. Then, the unit recognizer is embedded in the error analysis module. The digital information is input into the unit recognition channel corresponding to the bit information to obtain the unit abnormality parameter, thereby improving the key abnormality recognition accuracy of the digital cash register. The error analysis module includes a unit recognizer. The unit recognizer includes multiple unit recognition channels corresponding to multiple sample bit information. Each unit recognition channel includes an input layer, a hidden layer, and an output layer. Each unit recognition channel has the function of matching the unit abnormality parameter of the input digital information.

[0040] According to the position information, determine whether the user has previously input previous digital information, if so, extract and obtain at least one previous digital information and at least one previous position information, if not, calculate the comprehensive abnormality parameter according to the unit abnormality parameter;

[0041] According to the number of bits in the bit information, determine whether the current input has "previous digital information". If the number of bits is the first number of the current input, then the current input does not have "previous digital information", and the continuous abnormality parameter is recorded as 0. The unit abnormality parameter and the continuous abnormality parameter are weighted and calculated according to the comprehensive abnormality calculation formula to obtain the comprehensive abnormality parameter. The comprehensive abnormality calculation formula is:

[0042] ;

[0043] Among them, R is the output comprehensive abnormal parameter, X is the input unit abnormal parameter, and Y is the input continuous abnormal parameter. They are respectively a unit abnormality weight value and a continuous abnormality weight value pre-set and determined by the force feedback control system based on the digital cash register buttons.

[0044] Based on the historical input data of the digital cash register or the user, a continuous identifier is trained to identify at least one preceding digital information, at least one preceding position information, digital information and position information to obtain continuous abnormal parameters;

[0045] obtaining a plurality of sample continuous inputs according to historical input data of the digital cash register or the user in the at least one preceding position information within a historical time;

[0046] Collecting a plurality of sample input digital information sets in the bit information within a plurality of preset historical time periods under the continuous input of the plurality of samples;

[0047] Calculate and obtain multiple sample input probability sets of multiple sample digital information according to multiple sample input digital information sets;

[0048] Calculating and obtaining multiple sample continuous abnormal parameter sets according to the multiple sample input probability sets;

[0049] If the number code indicates that the currently input specific number is not the first number, then at least one preceding number information and at least one preceding position information are obtained. For example, when the number code indicates that the currently input specific number is the second number, the specific number corresponding to the first number is recorded as the preceding number information, and the digit and number code corresponding to the first number are recorded as the preceding position information.

[0050] The historical time includes a plurality of preset historical time periods. The historical input data of the digital cash register or the user in at least one previous digit information within the historical time is read to obtain a plurality of sample continuous inputs. For example, when a previous digit information is obtained, the plurality of sample continuous inputs include a plurality of historical digits corresponding to the tenth digit in the historical input data of the digital cash register or the user within the historical time. The plurality of historical digits may be 1, 2, 3, 4, etc.

[0051] Furthermore, multiple sets of sample input digital information in the bit information are collected within multiple preset historical time periods under multiple sample continuous inputs. Each set of sample input digital information includes multiple sample digital information of the bit information in multiple historical input data corresponding to each sample continuous input. For example, when the sample continuous input is 1, the corresponding multiple historical input data include 11, 11, 15, etc., and the bit information indicates that the specific number currently input is the second number, then the multiple sample digital information corresponding to the sample continuous input includes 1, 1, 5.

[0052] Further, multiple sample input probability sets are calculated based on multiple sample input digital information sets. Each sample input probability set includes multiple sample input probabilities corresponding to multiple sample digital information in each sample input digital information set. Each sample input probability includes the ratio of the total number of occurrences of each sample digital information in each sample input digital information set to the total number of multiple sample digital information.

[0053] Further, multiple sample continuous anomaly parameter sets are calculated based on multiple sample input probability sets. Each sample continuous anomaly parameter set includes multiple sample continuous anomaly parameters corresponding to multiple sample input probabilities in each sample input probability set. Each sample continuous anomaly parameter is the difference between 1 and the sample input probability.

[0054] Training a first continuous recognition channel in a continuous recognizer according to a first sample continuous input, a plurality of sample digital information and a first sample continuous abnormal parameter set in the plurality of sample continuous inputs and the plurality of sample continuous abnormal parameter sets;

[0055] Based on machine learning, a first continuous recognition channel corresponding to the first sample continuous input in the continuous recognizer is constructed;

[0056] Using multiple sample digital information and a first sample continuous abnormal parameter set, a first continuous recognition channel is trained once, and a first accuracy rate is obtained by testing, wherein each group of training data is trained for the first number of times during the training;

[0057] According to the deviation between the first accuracy rate and the convergence accuracy rate, adjusting and rounding the first number to obtain a second number;

[0058] Using multiple sample digital information and the first sample continuous abnormal parameter set, according to the second number, the first continuous recognition channel is trained twice, and the second accuracy is obtained by testing;

[0059] The training is continued until the accuracy of the first continuous recognition channel reaches a convergence accuracy.

[0060] A plurality of sample continuous inputs are randomly selected to obtain a first sample continuous input. The first sample continuous input may be any one of the plurality of sample continuous inputs. Then, the sample continuous abnormal parameter set corresponding to the first sample continuous input is recorded as the first sample continuous abnormal parameter set. The fully connected neural network is set as the basic network of the first continuous identification channel corresponding to the first sample continuous input. The fully connected neural network is an artificial neural network structure with a relatively simple connection method. The fully connected neural network includes an input layer, a hidden layer, and an output layer. Furthermore, the basic network is trained once according to the plurality of sample digital information corresponding to the first sample continuous input and the first sample continuous abnormal parameter set to obtain a first accuracy rate. Moreover, during one training, each set of training data is trained for the first number of times. The first accuracy rate includes the output accuracy rate parameter of the basic network during one training. Each set of training data includes a random sample digital information corresponding to the first sample continuous input, and the sample continuous abnormal parameter corresponding to the sample digital information. The first number of times includes the preset number of training times pre-set and determined by the force feedback control system based on the digital cash register key.

[0061] Further, the convergence accuracy rate includes an output accuracy rate threshold of the first continuous recognition channel pre-set and determined by the force feedback control system based on the digital cash register button. Determine whether the first accuracy rate reaches the convergence accuracy rate. If the first accuracy rate is greater than / equal to the convergence accuracy rate, then the first accuracy rate reaches the convergence accuracy rate and a first continuous recognition channel is generated. If the first accuracy rate is less than the convergence accuracy rate, then the difference between the convergence accuracy rate and the first accuracy rate is recorded as the deviation between the first accuracy rate and the convergence accuracy rate, and the first number is adjusted and rounded according to the deviation between the first accuracy rate and the convergence accuracy rate to obtain a second number. For example, the first accuracy rate is multiplied by the deviation between the first accuracy rate and the convergence accuracy rate and the first number to obtain the first compensation number. The sum of the first compensation number and the first number is output as the second number. Moreover, when the first compensation number is a non-integer, the first compensation number is rounded off and then the rounded first compensation number is added to the first number.

[0062] Further, each set of training data is trained twice according to the second number, and the second accuracy is obtained by testing. The second accuracy includes the output accuracy parameter of the basic network during the second training. Similarly, it is determined whether the second accuracy reaches the convergence accuracy. If the second accuracy is greater than / equal to the convergence accuracy, then the second accuracy reaches the convergence accuracy and a first continuous recognition channel is generated. If the second accuracy is less than the convergence accuracy, training is continued until the convergence accuracy is reached, and a first continuous recognition channel is generated. The first continuous recognition channel includes an input layer, a hidden layer, and an output layer.

[0063] Continue training to obtain multiple continuous recognition channels corresponding to other multiple sample continuous inputs to obtain the continuous recognizer;

[0064] A previous continuous input is generated according to at least one previous digital information and at least one previous bit information, and the current input is identified using a continuous identification channel corresponding to the previous continuous input of the continuous identifier to obtain continuous abnormality parameters.

[0065] According to the continuous abnormal parameters and the unit abnormal parameters, a comprehensive abnormal parameter is obtained by weighted calculation;

[0066] Continue to train multiple continuous recognition channels corresponding to other multiple sample continuous inputs, obtain the continuous recognition channel corresponding to each sample continuous input, and add the continuous recognition channel corresponding to each sample continuous input to the continuous recognizer. The multiple continuous recognition channels are constructed in the same way as the first continuous recognition channel. The continuous recognizer includes a continuous recognition channel corresponding to each sample continuous input. Then, at least one previous digital information and at least one previous bit information are recorded as the previous continuous input. The current input is used as input information and input into the continuous recognition channel corresponding to the previous continuous input to obtain continuous anomaly parameters, and the unit anomaly parameters and continuous anomaly parameters are weightedly calculated according to the comprehensive anomaly calculation formula to obtain comprehensive anomaly parameters.

[0067] Classify the comprehensive abnormal parameters into input abnormality levels to obtain input abnormality levels and force feedback levels;

[0068] As attached Figure 2 As shown, the comprehensive abnormal parameters are input into the abnormal level classification, which also includes:

[0069] Process and calculate to obtain a sample comprehensive abnormal parameter set according to the historical input data of the digital cash register or the user;

[0070] Clustering the sample comprehensive abnormal parameter set to obtain multiple sample clustering results, and setting multiple sample abnormality levels and multiple sample force feedback levels respectively;

[0071] Based on the comprehensive abnormal parameters of samples in the clustering results of multiple samples, the abnormal levels of multiple samples and the force feedback levels of multiple samples, a decision tree is used to construct a force feedback classifier;

[0072] The force feedback classifier is used to classify the comprehensive abnormal parameters to obtain the abnormality level and the force feedback level.

[0073] Through the force feedback control module, key force feedback is performed according to the force feedback level through the feedback component on the key component corresponding to the digital information.

[0074] The historical input data of the digital cash register or the user is calculated according to the comprehensive anomaly calculation formula to obtain a sample comprehensive anomaly parameter set. The sample comprehensive anomaly parameter set includes multiple historical comprehensive anomaly parameters. The calculation method of multiple historical comprehensive anomaly parameters is the same as that of the comprehensive anomaly parameters. Then, the sample comprehensive anomaly parameter set is clustered, and the same multiple historical comprehensive anomaly parameters are classified into one category to obtain multiple sample clustering results. Each sample clustering result includes the same multiple historical comprehensive anomaly parameters in the sample comprehensive anomaly parameter set.

[0075] According to multiple sample clustering results, multiple sample abnormality levels and multiple sample force feedback levels are set respectively. Each sample clustering result corresponds to a sample abnormality level and a sample force feedback level. Exemplarily, when setting multiple sample abnormality levels and multiple sample force feedback levels, it is judged whether the historical comprehensive abnormality parameter in each sample clustering result is less than the preset comprehensive abnormality parameter pre-set and determined by the force feedback control system based on the digital cash register button. If the historical comprehensive abnormality parameter in the sample clustering result is greater than the preset comprehensive abnormality parameter, then the sample force feedback level corresponding to the sample clustering result is recorded as a strong feedback level, and the ratio of the historical comprehensive abnormality parameter in the sample clustering result to the preset comprehensive abnormality parameter is set as the sample abnormality level. If the historical comprehensive abnormality parameter in the sample clustering result is less than / equal to the preset comprehensive abnormality parameter, then the sample force feedback level corresponding to the sample clustering result is recorded as a weak feedback level, and the ratio of the historical comprehensive abnormality parameter in the sample clustering result to the preset comprehensive abnormality parameter is set as the sample abnormality level. The strong feedback level and the weak feedback level both have corresponding marked vibration amplitudes and vibration durations. Moreover, the stronger the feedback level, the larger the corresponding vibration amplitude and the longer the vibration duration.

[0076] Further, according to the sample comprehensive abnormal parameters, multiple sample abnormal levels and multiple sample force feedback levels in the multiple sample clustering results, a decision tree is used to construct a force feedback classifier, and the force feedback classifier is embedded in the force feedback control module. Decision tree is a machine learning method. Decision tree is a tree structure. In the decision tree, each internal node represents a judgment on an attribute, each branch represents the output of a judgment result, and finally each leaf node represents a classification result. The force feedback classifier is a decision tree composed of sample comprehensive abnormal parameters, multiple sample abnormal levels and multiple sample force feedback levels in multiple sample clustering results. The force feedback classifier includes multiple force feedback classification nodes. Each force feedback classification node includes a sample comprehensive abnormal parameter in a random sample clustering result, and a sample abnormal level and a sample force feedback level corresponding to the sample clustering result. Then, the comprehensive abnormal parameter is input into the force feedback classifier, and the comprehensive abnormal parameter is classified by the multiple force feedback classification nodes in the force feedback classifier to obtain the abnormal level and the force feedback level. Multiple feedback components are arranged on multiple key components in the digital cash register. Each feedback component includes a vibration motor arranged on each key component in the digital cash register. According to the force feedback level, the feedback component on the key component corresponding to the digital information in the force feedback control module is subjected to key force feedback, that is, according to the vibration amplitude and vibration duration corresponding to the force feedback level, the vibration motor on the key component corresponding to the digital information is subjected to vibration control, thereby improving the force feedback intelligence of the keys of the digital cash register and improving the force feedback effect of the keys of the digital cash register.

[0077] In summary, the force feedback control method based on digital cash register buttons provided by the present application has the following technical effects:

[0078] The digital information and bit information currently input by the user through the digital cash register keys are collected through the input acquisition module; the digital information and bit information are identified through the error analysis module to obtain the unit abnormality parameter; it is determined whether the user has previously input the previous digital information, if so, at least one previous digital information and at least one previous bit information are extracted, if not, the comprehensive abnormality parameter is calculated according to the unit abnormality parameter; the continuous identifier is trained to identify at least one previous digital information, at least one previous bit information, digital information and bit information to obtain the continuous abnormality parameter; the comprehensive abnormality parameter is obtained by weighted calculation according to the continuous abnormality parameter and the unit abnormality parameter; the comprehensive abnormality parameter is classified into the input abnormality level to obtain the input abnormality level and the force feedback level; the feedback component on the key component corresponding to the digital information is fed back according to the force feedback level. The technical effect of improving the key abnormality recognition accuracy of the digital cash register, improving the key force feedback effect of the digital cash register, and improving the key force feedback intelligence of the digital cash register is achieved.

[0079] Embodiment 2

[0080] Based on the force feedback control method based on the digital cash register keys in the aforementioned embodiment, the present invention also provides a force feedback control system based on the digital cash register keys, see the attached Figure 3 The system is in communication with a force feedback control device based on a digital cash register key, the device comprising an input acquisition module, an error analysis module, a force feedback control module and a plurality of feedback components, the plurality of feedback components being arranged on a plurality of key components in the digital cash register, the system comprising:

[0081] A current input acquisition module, which is used to collect digital information and bit information currently input by the user through the keys of the digital cash register through the input acquisition module as the current input;

[0082] A unit abnormality recognition module, the unit abnormality recognition module is used to train a unit identifier based on the historical input data of the digital cash register or the user through the error analysis module, identify the digital information and the bit information, and obtain a unit abnormality parameter;

[0083] A previous digit judgment module, the previous digit judgment module is used to judge whether the user has previously input previous digital information according to the digit information, if so, extract and obtain at least one previous digital information and at least one previous digit information, if not, calculate the comprehensive abnormality parameter according to the unit abnormality parameter;

[0084] A continuous abnormality recognition module, the continuous abnormality recognition module is used to train a continuous recognizer based on the historical input data of the digital cash register or the user, recognize at least one previous digital information, at least one previous position information, digital information and position information, and obtain continuous abnormality parameters;

[0085] A comprehensive abnormality calculation module, the comprehensive abnormality calculation module is used to obtain a comprehensive abnormality parameter by weighted calculation according to the continuous abnormality parameter and the unit abnormality parameter;

[0086] An abnormality level classification module, the abnormality level classification module is used to classify the input abnormality level of the comprehensive abnormality parameters to obtain the input abnormality level and the force feedback level;

[0087] A force feedback module is used to perform key force feedback according to the force feedback level through a feedback component on a key component corresponding to the digital information through a force feedback control module.

[0088] Furthermore, the system also includes:

[0089] A user record acquisition module, the user record acquisition module is used to acquire user records based on login records in the digital cash register;

[0090] a first execution module, the first execution module being used to determine whether there is more than one user in the user record, and if not, training the unit identifier according to the historical input data in the digital cash register;

[0091] a second execution module, wherein if yes, the second execution module is used to obtain login time records of multiple users, and determine whether the login time difference between the maximum login time and the minimum login time is greater than a login time difference threshold according to the login time records;

[0092] The third execution module is used to train the unit identifier by using the historical input data of the digital cash register if yes, and train the unit identifier according to the historical input data of the user if no.

[0093] Furthermore, the system also includes:

[0094] A sample input digital information set acquisition module, the sample input digital information set acquisition module is used to acquire multiple sample input digital information sets of multiple sample bit information within multiple preset historical time periods according to the historical input data of the digital cash register or the user;

[0095] A sample occurrence probability set acquisition module, the sample occurrence probability set acquisition module is used to calculate the probability of occurrence of multiple sample digital information in multiple sample bit information according to multiple sample input digital information sets, and obtain multiple sample occurrence probability sets;

[0096] A fourth execution module, the fourth execution module is used to calculate and obtain multiple sample unit abnormal parameter sets according to multiple sample occurrence probability sets;

[0097] a fifth execution module, the fifth execution module being used to respectively use a plurality of sample digital information and a plurality of sample unit abnormal parameter sets to train the unit identifier, the unit identifier comprising a plurality of unit identification channels corresponding to the plurality of sample bit information;

[0098] A unit abnormality parameter determination module is used to use the unit identification channel corresponding to the bit information in the unit identifier to identify the digital information and obtain the unit abnormality parameter.

[0099] Furthermore, the system also includes:

[0100] A unit identification channel construction module, wherein the unit identification channel construction module is used to respectively construct a plurality of unit identification channels corresponding to a plurality of sample bit information based on machine learning;

[0101] A channel training module, wherein the channel training module is used to take digital information as input and unit abnormality parameters as output, and respectively use the multiple sample digital information and the multiple sample unit abnormality parameter sets to train multiple unit recognition channels until convergence;

[0102] A channel integration module is used to integrate the multiple unit identification channels to obtain the unit identifier.

[0103] Furthermore, the system also includes:

[0104] A sample continuous input obtaining module, the sample continuous input obtaining module is used to obtain a plurality of sample continuous inputs according to the historical input data of the digital cash register or the user in the at least one preceding position information within a historical time;

[0105] A sample input digital information set acquisition module, the sample input digital information set acquisition module is used to collect multiple sample input digital information sets in the bit information within multiple preset historical time periods under the continuous input of the multiple samples;

[0106] A sample input probability calculation module, the sample input probability calculation module is used to calculate multiple sample input probability sets of multiple sample digital information according to multiple sample input digital information sets;

[0107] A sample continuous abnormal parameter calculation module, the sample continuous abnormal parameter calculation module is used to calculate and obtain multiple sample continuous abnormal parameter sets according to the multiple sample input probability sets;

[0108] a sixth execution module, the sixth execution module being used to train a first continuous recognition channel in a continuous identifier according to a first sample continuous input, a plurality of sample digital information and a first sample continuous abnormal parameter set in the plurality of sample continuous inputs and the plurality of sample continuous abnormal parameter sets;

[0109] A continuous identifier acquisition module, the continuous identifier acquisition module is used to continue training to obtain multiple continuous recognition channels corresponding to other multiple sample continuous inputs to obtain the continuous identifier;

[0110] A continuous abnormal parameter acquisition module is used to generate a previous continuous input based on at least one previous digital information and at least one previous bit information, and use the continuous recognition channel corresponding to the previous continuous input of the continuous identifier to identify the current input to obtain continuous abnormal parameters.

[0111] Furthermore, the system also includes:

[0112] A seventh execution module, the seventh execution module is used to construct a first continuous recognition channel corresponding to the continuous input of the first sample in the continuous recognizer based on machine learning;

[0113] A first accuracy acquisition module, the first accuracy acquisition module is used to use multiple sample digital information and a first sample continuous abnormal parameter set to train a first continuous recognition channel once, and test to obtain a first accuracy, in which each group of training data is trained for the first number of times;

[0114] A second number obtaining module, wherein the second number obtaining module is used to adjust and calculate the first number and round it according to the deviation between the first accuracy rate and the convergence accuracy rate to obtain a second number;

[0115] A second accuracy acquisition module, wherein the second accuracy acquisition module is used to use a plurality of sample digital information and a first sample continuous abnormal parameter set, perform a second training on the first continuous recognition channel according to a second number, and test to obtain a second accuracy;

[0116] An eighth execution module, wherein the eighth execution module is used to continue training until the accuracy of the first continuous recognition channel reaches a convergence accuracy.

[0117] Furthermore, the system also includes:

[0118] A sample comprehensive abnormal parameter set acquisition module, the sample comprehensive abnormal parameter set acquisition module is used to process and calculate to obtain a sample comprehensive abnormal parameter set according to the historical input data of the digital cash register or the user;

[0119] A sample clustering module, the sample clustering module is used to cluster the sample comprehensive abnormal parameter set to obtain multiple sample clustering results, and respectively set multiple sample abnormality levels and multiple sample force feedback levels;

[0120] A force feedback classifier construction module, wherein the force feedback classifier construction module is used to construct a force feedback classifier using a decision tree based on a sample comprehensive abnormality parameter, a plurality of sample abnormality levels and a plurality of sample force feedback levels in a plurality of sample clustering results;

[0121] A comprehensive abnormality classification module is used to use the force feedback classifier to classify the comprehensive abnormality parameters to obtain the abnormality level and the force feedback level.

[0122] A force feedback control system based on digital cash register buttons provided in an embodiment of the present invention can execute a force feedback control method based on digital cash register buttons provided in any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.

[0123] The modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of the functional modules are only for the convenience of distinguishing each other and are not used to limit the scope of protection of the present invention.

[0124] The present application provides a force feedback control method based on digital cash register keys, wherein the method is applied to a force feedback control system based on digital cash register keys, and the method includes: collecting digital information and bit information currently input by the user through the digital cash register keys through an input acquisition module; identifying the digital information and bit information through an error analysis module to obtain a unit abnormality parameter; judging whether the user has input the previous digital information, if so, extracting and obtaining at least one previous digital information and at least one previous bit information, if not, calculating the comprehensive abnormality parameter according to the unit abnormality parameter; training a continuous identifier to identify at least one previous digital information, at least one previous bit information, digital information and bit information to obtain a continuous abnormality parameter; weighted calculation to obtain a comprehensive abnormality parameter according to the continuous abnormality parameter and the unit abnormality parameter; classifying the comprehensive abnormality parameter into an input abnormality level to obtain an input abnormality level and a force feedback level; and performing key force feedback on the feedback component on the key component corresponding to the digital information according to the force feedback level. The present application solves the technical problems of low key abnormality recognition accuracy of digital cash registers and poor key force feedback effect of digital cash registers in the prior art. The technical effects of improving the abnormal recognition accuracy of key presses of a digital cash register, improving the key force feedback effect of the digital cash register, and improving the intelligence of the key force feedback of the digital cash register are achieved.

[0125] Although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention. The scope of the present invention is determined by the scope of the appended claims.

Claims

1. A force feedback control method based on digital cash register buttons, characterized in that: The method is applied to a force feedback control device based on a digital cash register key, the device comprises an input acquisition module, an error analysis module, a force feedback control module and a plurality of feedback components, the plurality of feedback components are arranged on a plurality of key components in the digital cash register, the method comprises: The input collection module collects the digital information and bit information currently input by the user through the keys of the digital cash register as the current input; Through the error analysis module, based on the historical input data of the digital cash register or the user, a unit identifier is trained to identify the digital information and the bit information to obtain unit abnormality parameters; According to the position information, determine whether the user has previously input previous digital information, if so, extract and obtain at least one previous digital information and at least one previous position information, if not, calculate the comprehensive abnormality parameter according to the unit abnormality parameter; Based on the historical input data of the digital cash register or the user, a continuous identifier is trained to identify at least one preceding digital information, at least one preceding position information, digital information and position information to obtain continuous abnormal parameters; According to the continuous abnormal parameters and the unit abnormal parameters, a comprehensive abnormal parameter is obtained by weighted calculation; Classify the comprehensive abnormal parameters into input abnormality levels to obtain input abnormality levels and force feedback levels; By means of a force feedback control module, a feedback component on a key component corresponding to the digital information is used to perform key force feedback according to the force feedback level; Based on the login record in the digital cash register, obtaining the user record; determining whether there is more than one user in the user record, and if not, training the unit identifier based on historical input data in the digital cash register; If yes, then obtain the login time records of multiple users, and determine whether the login time difference between the maximum login time and the minimum login time is greater than the login time difference threshold according to the login time records; If yes, the unit identifier is trained using the historical input data of the digital cash register, and if no, the unit identifier is trained based on the historical input data of the user; According to the historical input data of the digital cash register or the user, a plurality of sample input digital information sets of a plurality of sample bit information within a plurality of preset historical time periods are obtained; According to the multiple sample input digital information sets, the probability of occurrence of the multiple sample digital information in the multiple sample bit information is calculated to obtain the multiple sample occurrence probability sets; According to the occurrence probability sets of multiple samples, multiple sample unit abnormal parameter sets are calculated and obtained; Using a plurality of sample digital information and a plurality of sample unit abnormal parameter sets respectively, the unit identifier is trained, wherein the unit identifier includes a plurality of unit identification channels corresponding to the plurality of sample bit information; The unit identification channel corresponding to the bit information in the unit identifier is used to identify the digital information to obtain the unit abnormality parameter.

2. The method according to claim 1, characterized in that The method comprises: Based on machine learning, multiple unit recognition channels corresponding to multiple sample bit information are constructed respectively; Taking digital information as input and unit abnormality parameters as output, respectively using the plurality of sample digital information and the plurality of sample unit abnormality parameter sets to train the plurality of unit recognition channels until convergence; The plurality of unit identification channels are integrated to obtain the unit identifier.

3. The method according to claim 1, characterized in that: The method comprises: obtaining a plurality of sample continuous inputs according to historical input data of the digital cash register or the user in the at least one preceding position information within a historical time; Collecting a plurality of sample input digital information sets in the bit information within a plurality of preset historical time periods under the continuous input of the plurality of samples; Calculate and obtain multiple sample input probability sets of multiple sample digital information according to multiple sample input digital information sets; Calculating and obtaining multiple sample continuous abnormal parameter sets according to the multiple sample input probability sets; Training a first continuous recognition channel in a continuous recognizer according to a first sample continuous input, a plurality of sample digital information and a first sample continuous abnormal parameter set in the plurality of sample continuous inputs and the plurality of sample continuous abnormal parameter sets; Continue training to obtain multiple continuous recognition channels corresponding to other multiple sample continuous inputs to obtain the continuous recognizer; A previous continuous input is generated according to at least one previous digital information and at least one previous bit information, and the current input is identified using a continuous identification channel corresponding to the previous continuous input of the continuous identifier to obtain continuous abnormality parameters.

4. The method according to claim 3, characterized in that The method comprises: Based on machine learning, a first continuous recognition channel corresponding to the first sample continuous input in the continuous recognizer is constructed; Using multiple sample digital information and a first sample continuous abnormal parameter set, a first continuous identification channel is trained once, and a first accuracy rate is obtained by testing, wherein each group of training data is trained for a first number of times during the training; According to the deviation between the first accuracy rate and the convergence accuracy rate, adjusting and rounding the first number to obtain a second number; Using multiple sample digital information and the first sample continuous abnormal parameter set, according to the second number, the first continuous recognition channel is trained twice, and the second accuracy is obtained by testing; The training is continued until the accuracy of the first continuous recognition channel reaches a convergence accuracy.

5. The method according to claim 1, characterized in that The method comprises: Process and calculate to obtain a sample comprehensive abnormal parameter set according to the historical input data of the digital cash register or the user; Clustering the sample comprehensive abnormal parameter set to obtain multiple sample clustering results, and setting multiple sample abnormality levels and multiple sample force feedback levels respectively; Based on the comprehensive abnormal parameters of samples in the clustering results of multiple samples, the abnormal levels of multiple samples and the force feedback levels of multiple samples, a decision tree is used to construct a force feedback classifier; The force feedback classifier is used to classify the comprehensive abnormal parameters to obtain the abnormality level and the force feedback level.

6. A force feedback control system based on digital cash register buttons, characterized in that: The system is used to execute the method according to any one of claims 1 to 5, the system is communicatively connected with a force feedback control device based on a key of a digital cash register, the device comprises an input acquisition module, an error analysis module, a force feedback control module and a plurality of feedback components, the plurality of feedback components are arranged on a plurality of key components in the digital cash register, the system comprises: A current input acquisition module, which is used to collect digital information and bit information currently input by the user through the keys of the digital cash register through the input acquisition module as the current input; A unit abnormality recognition module, the unit abnormality recognition module is used to train a unit identifier based on the historical input data of the digital cash register or the user through the error analysis module, identify the digital information and the bit information, and obtain a unit abnormality parameter; A previous digit judgment module, the previous digit judgment module is used to judge whether the user has previously input previous digital information according to the digit information, if so, extract and obtain at least one previous digital information and at least one previous digit information, if not, calculate the comprehensive abnormality parameter according to the unit abnormality parameter; A continuous abnormality recognition module, the continuous abnormality recognition module is used to train a continuous recognizer based on the historical input data of the digital cash register or the user, recognize at least one previous digital information, at least one previous position information, digital information and position information, and obtain continuous abnormality parameters; A comprehensive abnormality calculation module, the comprehensive abnormality calculation module is used to obtain a comprehensive abnormality parameter by weighted calculation according to the continuous abnormality parameter and the unit abnormality parameter; An abnormality level classification module, the abnormality level classification module is used to classify the input abnormality level of the comprehensive abnormality parameters to obtain the input abnormality level and the force feedback level; A force feedback module, the force feedback module is used to perform key force feedback according to the force feedback level through the feedback component on the key component corresponding to the digital information through the force feedback control module; The system further comprises: A user record acquisition module, the user record acquisition module is used to acquire user records based on login records in the digital cash register; a first execution module, the first execution module being used to determine whether there is more than one user in the user record, and if not, training the unit identifier according to the historical input data in the digital cash register; a second execution module, wherein if yes, the second execution module is used to obtain login time records of multiple users, and determine whether the login time difference between the maximum login time and the minimum login time is greater than a login time difference threshold according to the login time records; a third execution module, the third execution module being used to, if yes, train the unit identifier using the historical input data of the digital cash register, and if no, train the unit identifier according to the historical input data of the user; A sample input digital information set acquisition module, the sample input digital information set acquisition module is used to acquire multiple sample input digital information sets of multiple sample bit information within multiple preset historical time periods according to the historical input data of the digital cash register or the user; A sample occurrence probability set acquisition module, the sample occurrence probability set acquisition module is used to calculate the probability of occurrence of multiple sample digital information in multiple sample bit information according to multiple sample input digital information sets, and obtain multiple sample occurrence probability sets; A fourth execution module, the fourth execution module is used to calculate and obtain multiple sample unit abnormal parameter sets according to multiple sample occurrence probability sets; a fifth execution module, the fifth execution module being used to respectively use a plurality of sample digital information and a plurality of sample unit abnormal parameter sets to train the unit identifier, the unit identifier comprising a plurality of unit identification channels corresponding to the plurality of sample bit information; A unit abnormality parameter determination module is used to use the unit identification channel corresponding to the bit information in the unit identifier to identify the digital information and obtain the unit abnormality parameter.

Images