A drop height adjustment system for a drop dispenser

Abstract

The application provides a drop height adjusting system for a drop-type vending machine, which comprises a first determining module, a first detecting module, a moving module, a second detecting module and a control module. The control module is connected with the first detecting module, the moving module and the second detecting module respectively, receives distance information sent by the first detecting module, judges whether the distance information is greater than a preset distance, controls the moving module to move to a drop level of a target product when it is determined that the distance information is greater than the preset distance, and controls the moving module to move to a delivery port when the moving module moves to the drop level and the second detecting module determines that the target product drops onto the moving module. The drop height of the delivery can be accurately adjusted, and the target product can be prevented from being damaged by dropping.

Description

Technical Field

[0001] This invention relates to the field of vending machine technology, and in particular to a dispensing drop height adjustment system for a drop-down vending machine. Background Technology

[0002] Currently, with the development of smart technology and e-commerce, vending machines still occupy an irreplaceable position in the market. Vending machines have made many innovations in payment methods, sales, and restocking, but in terms of dispensing, they still frequently encounter problems such as products falling and getting damaged, and jams. Even with offline intervention by administrators, there are issues such as excessive time delays and a significant decrease in user experience. Summary of the Invention

[0003] The present invention aims to at least partially solve one of the technical problems in the aforementioned technologies. Therefore, the object of the present invention is to provide a dispensing drop height adjustment system for a drop-type vending machine, accurately adjusting the drop height to reduce damage to goods due to falling.

[0004] To achieve the above objectives, embodiments of the present invention propose a dispensing drop height adjustment system for a drop-type vending machine, comprising:

[0005] The first determining module is used to receive the user's order information, parse the order information, and determine the target product;

[0006] The first detection module is used to detect the distance information between the target product and the shipping port;

[0007] Mobile module;

[0008] Second detection module;

[0009] The control module, connected to the first detection module, the movement module, and the second detection module respectively, is used for:

[0010] The system receives distance information sent by the first detection module and determines whether it is greater than a preset distance. When it is determined that the distance information is greater than the preset distance, the system controls the moving module to move to the horizontal plane where the target product will fall.

[0011] When the moving module moves to the falling horizontal plane and the second detection module determines that the target product has fallen onto the moving module, the moving module is controlled to move to the shipping port.

[0012] According to some embodiments of the present invention, the first determining module includes:

[0013] The first receiving module is used to receive the user's order information;

[0014] The parsing module is used for:

[0015] Keyword extraction is performed on the order information to determine target keywords;

[0016] The target keyword is matched with preset product names in the database to determine the preset product name with the highest matching degree with the target keyword, and the target product is determined based on the preset product name with the highest matching degree with the target keyword.

[0017] According to some embodiments of the present invention, the parsing module includes:

[0018] The conversion module is used to convert the order information to obtain the first text;

[0019] The preprocessing module is used to identify the type of characters included in the first text and remove characters that are consistent with the type included in the blacklist to obtain the second text;

[0020] The second determining module is used for:

[0021] Extract the data from the second text to obtain a dataset;

[0022] Based on the SVM classifier, the text information included in the dataset is classified to obtain several different types of statements;

[0023] The sentence is segmented using a word segmenter constructed from a dictionary, and the attribute information of each word in the sentence is determined.

[0024] Based on dependency syntax rules and the attribute information, a diagram illustrating the relationships between various words is determined;

[0025] The first filtering module is used for:

[0026] The association diagram is analyzed to determine the connection relationship between each word and other words. The word with the most connection relationships is selected as the target word.

[0027] The target words are converted into corresponding word vectors.

[0028] Calculate the distance between the word vector and the word vector of the preset standard word, and select the target words corresponding to the word vectors with a distance less than the preset distance as target keywords.

[0029] According to some embodiments of the present invention, the first detection module includes:

[0030] The acquisition module is used to acquire the distribution information of the storage shelves in the vending machine and generate a shelf distribution map based on the distribution information;

[0031] The tag module is used for:

[0032] Determine the storage shelf where the target product is located, and designate it as the target shelf;

[0033] Mark the target shelves in the shelf distribution map to obtain the marked points;

[0034] The third determining module is used to determine the distance information between the target product and the shipping outlet based on the location information of the marker point in the shelf distribution map, the location information of the shipping outlet, and the attribute information of the shelf distribution map.

[0035] According to some embodiments of the present invention, a moving module includes:

[0036] The second receiving module is used to receive instructions sent by the control module;

[0037] Mobile frame;

[0038] A drive module is used to drive the moving frame to move up or down according to the instructions.

[0039] According to some embodiments of the present invention, the second detection module includes a pressure sensor.

[0040] According to some embodiments of the present invention, a third detection module is used to detect whether the drive module is faulty;

[0041] An alarm module is used to issue an alarm prompt when the third detection module determines that the drive module has malfunctioned.

[0042] According to some embodiments of the present invention, the third detection module includes:

[0043] The first acquisition module is used to acquire the sound signal generated by the vibration of the drive module;

[0044] The judgment module is used for:

[0045] The sound signal is analyzed to obtain sound features;

[0046] The sound feature is compared with the standard sound feature corresponding to the standard sound signal. If the difference between the sound feature and the standard sound feature is greater than a preset difference value, it indicates that the driving module has malfunctioned.

[0047] According to some embodiments of the present invention, the first determining module includes:

[0048] The second acquisition module is used to acquire the user's voice signal;

[0049] The fourth determining module is used for:

[0050] The speech signal is segmented into frames to obtain sub-speech signals of several fixed length frames; the sub-speech signals are subjected to Fourier transform to obtain frequency domain signals, and the spectrum is determined based on the frequency domain signals.

[0051] Determine the sound wave frequency in the spectrum diagram and determine whether it is within the preset sound wave frequency range;

[0052] Determine the sound wave intensity in the spectrum and determine whether it is within a preset sound wave intensity range;

[0053] Spectrum diagrams with sound wave frequencies and intensities outside the preset range are selected as abnormal spectrum diagrams; the sub-speech signals corresponding to the abnormal spectrum diagrams are taken as noisy signals; the noise intensity characteristics and distribution characteristics of each noisy signal are determined.

[0054] All spectrograms except those with abnormal spectrograms are considered normal spectrograms; the sub-speech signals corresponding to the normal spectrograms are considered clean signals.

[0055] The fifth determining module is used for:

[0056] The noise intensity and distribution features are input into a pre-trained noise recognition model to determine the target noise scene;

[0057] Noise reduction module, used for:

[0058] According to the target noise scene, a preset noise scene-noise reduction coefficient table is queried to determine the corresponding noise reduction coefficient. The noisy signal is then denoised according to the noise reduction coefficient to obtain a denoised signal.

[0059] The reconstruction module is used to: reconstruct the signal based on the noise-reduced signal and the clean signal to obtain the target signal;

[0060] The recognition module is used for:

[0061] Perform speech recognition on the target signal and determine the speech recognition result;

[0062] The user's order information is determined based on the speech recognition results.

[0063] According to some embodiments of the present invention, the identification module includes:

[0064] The correction module is used to determine the signal waveform curve corresponding to the target signal, delete the blank segments at both ends of the signal waveform curve included in the signal waveform curve, mark the speech pause position and time, and obtain the corrected signal waveform curve.

[0065] An extraction module is used to extract speech feature sequences from the corrected signal waveform curves included in the corrected signal waveform curve diagram;

[0066] The matching module is used to learn the speech feature sequence based on a preset learning algorithm to obtain phoneme information; and to match the phoneme information with a preset acoustic model to obtain speech recognition results.

[0067] The conversion module is used to convert the speech recognition results into text based on a language model to determine the user's order information.

[0068] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0069] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0070] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0071] Figure 1 This is a block diagram of a dispensing drop height adjustment system for a drop-type vending machine according to an embodiment of the present invention.

[0072] Figure 2 This is a block diagram of a first determining module according to an embodiment of the present invention;

[0073] Figure 3 This is a block diagram of a parsing module according to an embodiment of the present invention. Detailed Implementation

[0074] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0075] like Figure 1 As shown in the figure, an embodiment of the present invention proposes a dispensing drop height adjustment system for a drop-type vending machine, comprising:

[0076] The first determining module is used to receive the user's order information, parse the order information, and determine the target product;

[0077] The first detection module is used to detect the distance information between the target product and the shipping port;

[0078] Mobile module;

[0079] Second detection module;

[0080] The control module, connected to the first detection module, the movement module, and the second detection module respectively, is used for:

[0081] The system receives distance information sent by the first detection module and determines whether it is greater than a preset distance. When it is determined that the distance information is greater than the preset distance, the system controls the moving module to move to the horizontal plane where the target product will fall.

[0082] When the moving module moves to the falling horizontal plane and the second detection module determines that the target product has fallen onto the moving module, the moving module is controlled to move to the shipping port.

[0083] The working principle of the above technical solution is as follows: The first determining module receives the user's order information, parses the order information, and determines the target product; the first detection module detects the distance information between the target product and the shipping port; the moving module receives the instructions sent by the control module and drives the moving frame to move up or down according to the instructions; the instructions include the moving direction and the moving distance. ； The control module receives distance information sent by the first detection module and determines whether it is greater than a preset distance. When the distance information is determined to be greater than the preset distance, the control module moves the mobile module to the drop surface of the target product. The preset distance is the distance that the product should maintain from the drop surface when the mobile module picks up the product. The second detection module determines whether the product has fallen. When the mobile module moves to the drop surface and the second detection module determines that the target product has fallen onto the mobile module, the control module moves the mobile module to the delivery port.

[0084] The beneficial effects of the above technical solution are as follows: The first detection module controls the moving device to move to the appropriate height to retrieve the product by detecting the drop distance of the target product, thus increasing the protection of the product during shipment. The third detection module can effectively determine whether the moving module has successfully retrieved the product, reducing the user's waiting time and improving the user experience.

[0085] According to some embodiments of the present invention, such as Figure 2 As shown, the first determining module includes:

[0086] The first receiving module is used to receive the user's order information;

[0087] The parsing module is used for:

[0088] Keyword extraction is performed on the order information to determine target keywords;

[0089] The target keyword is matched with preset product names in the database to determine the preset product name with the highest matching degree with the target keyword, and the target product is determined based on the preset product name with the highest matching degree with the target keyword.

[0090] The working principle of the above technical solution is as follows: The first receiving module is used to receive the user's order information, which includes product name and quantity information; the parsing module extracts keywords from the order information, determines the target keywords, and then matches the target keywords with preset product names in the database to determine the preset product name with the highest matching degree with the target keywords, and determines the target product based on the preset product name with the highest matching degree.

[0091] The beneficial effects of the above technical solution are as follows: Utilizing keyword extraction improves the accuracy of identifying target products. This effect is even more pronounced when the order information provided by the user is disorganized. Specifically, when the order information is disorganized, extracting keywords and then matching the extracted keywords with preset product names can quickly and effectively obtain useful information, thereby identifying the target product.

[0092] According to some embodiments of the present invention, such as Figure 3 As shown, the parsing module includes:

[0093] The conversion module is used to convert the order information to obtain the first text;

[0094] The preprocessing module is used to identify the type of characters included in the first text and remove characters that are consistent with the type included in the blacklist to obtain the second text;

[0095] The second determining module is used for:

[0096] Extract the data from the second text to obtain a dataset;

[0097] Based on the SVM classifier, the text information included in the dataset is classified to obtain several different types of statements;

[0098] The sentence is segmented using a word segmenter constructed from a dictionary, and the attribute information of each word in the sentence is determined.

[0099] Based on dependency syntax rules and the attribute information, a diagram illustrating the relationships between various words is determined;

[0100] The first filtering module is used for:

[0101] The association diagram is analyzed to determine the connection relationship between each word and other words. The word with the most connection relationships is selected as the target word.

[0102] The target words are converted into corresponding word vectors.

[0103] Calculate the distance between the word vector and the word vector of the preset standard word, and select the target words corresponding to the word vectors with a distance less than the preset distance as target keywords.

[0104] The working principle of the above technical solution is as follows: In the parsing module, the conversion module converts the order information to obtain the first text, which is the order product information parsed for the first time; the preprocessing module identifies the type of characters included in the first text and removes characters of the same type as those included in the blacklist to obtain the second text, which is a more concise and clear order product information after the invalid characters in the first text have been removed; the second determination module extracts the second text to obtain a data set, and then classifies the text information included in the data set based on an SVM classifier to obtain several different types of statements, according to a dictionary. The constructed word segmenter segments the statement, determines the attribute information of each word in the statement, and finally determines the relationship diagram between each word based on dependency syntax rules and the attribute information. The dictionary is an encyclopedia dictionary. The first filtering module first parses the relationship diagram to determine the connection relationship between each word and other words. The word with the most connection relationships is taken as the target word. Then, several target words are converted into corresponding word vectors. The distance between each word vector and the word vector of the preset standard word is calculated. The target words corresponding to the word vectors with a distance less than the preset distance are filtered out and taken as target keywords. The attribute information is the part of speech of the word, namely noun, verb, adjective, numeral, pronoun, preposition, auxiliary word, and conjunction.

[0105] The beneficial effects of the above technical solution are as follows: Preprocessing removes characters of the same type as those in the blacklist from the first text, resulting in cleaner text information and improving the efficiency and accuracy of keyword extraction. Dependency syntax rules and word attribute information more clearly demonstrate the relationships between words. The method of filtering keywords by calculating the distance between the word vector and the word vector of preset standard words allows for faster and more accurate keyword extraction, reducing user transaction waiting time and improving user experience.

[0106] According to some embodiments of the present invention, the first detection module includes:

[0107] The acquisition module is used to acquire the distribution information of the storage shelves in the vending machine and generate a shelf distribution map based on the distribution information;

[0108] The tag module is used for:

[0109] Determine the storage shelf where the target product is located, and designate it as the target shelf;

[0110] Mark the target shelves in the shelf distribution map to obtain the marked points;

[0111] The third determining module is used to determine the distance information between the target product and the shipping outlet based on the location information of the marker point in the shelf distribution map, the location information of the shipping outlet, and the attribute information of the shelf distribution map.

[0112] The working principle of the above technical solution is as follows: An acquisition module acquires the distribution information of the storage shelves in the vending machine and generates a shelf distribution map based on this information. This shelf distribution map shows the distribution of each product on different shelves within the vending machine. A marking module determines the storage shelf where the target product is located, designating it as the target shelf, and marks the target shelf on the shelf distribution map, obtaining a marker point. A third determining module determines the distance information between the target product and the dispensing outlet based on the location information of the marker point on the shelf distribution map, the location information of the dispensing outlet, and the attribute information of the shelf distribution map. The attribute information of the shelf distribution map includes the model and dimensions of the shelves.

[0113] The beneficial effects of the above technical solution are: accurate reference data can be obtained based on the shelf attribute information stored in the vending machine, and on this basis, the drop rate of the target product can be detected more accurately based on the reference data and the mark point of the target product.

[0114] According to some embodiments of the present invention, the mobile module includes:

[0115] The second receiving module is used to receive instructions sent by the control module;

[0116] Mobile frame;

[0117] A drive module is used to drive the moving frame to move up or down according to the instructions.

[0118] The working principle of the above technical solution is as follows: The second receiving module receives the instructions sent by the control module, namely the movement instructions; the driving module drives the moving frame to move upward or downward according to the instructions. The moving frame includes a water platform and foldable lifting supports connected to its two lower sides.

[0119] The beneficial effects of the above technical solution are: it facilitates the rapid up-and-down movement of the moving module after receiving instructions from the control module, thereby improving operating efficiency.

[0120] According to some embodiments of the present invention, the second detection module includes a pressure sensor.

[0121] The working principle of the above technical solution is as follows: The pressure sensor is used to detect the pressure data after receiving the goods.

[0122] The beneficial effects of the above technical solution are as follows: The pressure sensor effectively detects the product's dispensing status. When a user submits an order, the vending machine drives the moving module to retrieve the product. However, a malfunction may prevent the product from being dispensed. Adding a pressure sensor effectively determines whether the product is being dispensed correctly, improving the user experience.

[0123] According to some embodiments of the present invention, the mobile module includes:

[0124] The third detection module is used to detect whether the driver module is faulty.

[0125] An alarm module is used to issue an alarm prompt when the third detection module determines that the drive module has malfunctioned.

[0126] The working principle of the above technical solution is as follows: the third detection module detects whether the drive module is faulty; the alarm module issues an alarm prompt when the third detection module determines that the drive module has failed.

[0127] The beneficial effects of the above technical solution are as follows: Adding a third detection module facilitates the status monitoring of the vending machine. The drive module is a crucial part of the vending machine's dispensing process, therefore, the monitoring of its operational status is very important. The alarm module can promptly alert the administrator that the vending machine is malfunctioning and requires timely repair, and can also remind the user to use another vending machine, reducing user waiting time.

[0128] According to some embodiments of the present invention, the third detection module includes:

[0129] The first acquisition module is used to acquire the sound signal generated by the vibration of the drive module;

[0130] The judgment module is used for:

[0131] The sound signal is analyzed to obtain sound features;

[0132] The sound feature is compared with the standard sound feature corresponding to the standard sound signal. If the difference between the sound feature and the standard sound feature is greater than a preset difference value, it indicates that the driving module has malfunctioned.

[0133] The working principle of the above technical solution is as follows: The first acquisition module acquires the sound signal generated by the vibration of the drive module; the judgment module first analyzes the sound signal to obtain sound features, and then compares the sound features with the standard sound features corresponding to the standard sound signal. When the difference between the sound features and the standard sound features is greater than a preset difference value, it indicates that the drive module has malfunctioned. Specifically, the standard sound signal refers to the sound signal of the vending machine during normal operation preset by the system.

[0134] The beneficial effects of the above technical solution are: the method of detecting the working status of the drive module by comparing the sound signal generated by the vibration of the drive module with the standard sound signal can improve the accuracy of fault diagnosis.

[0135] According to some embodiments of the present invention, the first determining module includes:

[0136] The fourth determining module is used for:

[0137] The speech signal is segmented into frames to obtain sub-speech signals of several fixed length frames; the sub-speech signals are subjected to Fourier transform to obtain frequency domain signals, and the spectrum is determined based on the frequency domain signals.

[0138] Determine the sound wave frequency in the spectrum diagram and determine whether it is within the preset sound wave frequency range;

[0139] Determine the sound wave intensity in the spectrum and determine whether it is within a preset sound wave intensity range;

[0140] Spectrum diagrams with sound wave frequencies and intensities outside the preset range are selected as abnormal spectrum diagrams; the sub-speech signals corresponding to the abnormal spectrum diagrams are taken as noisy signals; the noise intensity characteristics and distribution characteristics of each noisy signal are determined.

[0141] All spectrograms except those with abnormal spectrograms are considered normal spectrograms; the sub-speech signals corresponding to the normal spectrograms are considered clean signals.

[0142] The fifth determining module is used for:

[0143] The noise intensity and distribution features are input into a pre-trained noise recognition model to determine the target noise scene;

[0144] Noise reduction module, used for:

[0145] According to the target noise scene, a preset noise scene-noise reduction coefficient table is queried to determine the corresponding noise reduction coefficient. The noisy signal is then denoised according to the noise reduction coefficient to obtain a denoised signal.

[0146] The reconstruction module is used to: reconstruct the signal based on the noise-reduced signal and the clean signal to obtain the target signal;

[0147] The recognition module is used for:

[0148] Perform speech recognition on the target signal and determine the speech recognition result;

[0149] The user's order information is determined based on the speech recognition results.

[0150] The working principle of the above technical solution is as follows: The second acquisition module acquires the user's voice signal; the fourth determination module performs frame-by-frame processing on the voice signal to obtain several fixed-length frames of sub-voice signals; Fourier transform processing is performed on the sub-voice signals to obtain frequency domain signals, and a spectrum diagram is determined based on the frequency domain signals; the sound wave frequency in the spectrum diagram is then determined to see if it is within a preset sound wave frequency range; the sound wave intensity in the spectrum diagram is determined to see if it is within a preset sound wave intensity range; spectrum diagrams whose sound wave frequency and intensity are both outside the preset sound wave frequency and intensity ranges are selected as abnormal spectrum diagrams; the sub-voice signals corresponding to the abnormal spectrum diagrams are taken as noisy signals; the noise intensity characteristics and distribution characteristics of each noisy signal are determined; the spectrum diagrams other than the abnormal spectrum diagrams are taken as normal spectrum diagrams; the normal spectrum diagrams are... The corresponding sub-speech signal is used as the clean signal; the fifth determining module inputs the noise intensity features and distribution features into a pre-trained noise recognition model to determine the target noise scene; the target noise scene, for example, different scenes such as schools, parks, and shopping malls; the noise reduction module queries a preset noise scene-noise reduction coefficient table according to the target noise scene to determine the corresponding noise reduction coefficient, and performs noise reduction processing on the noisy signal according to the noise reduction coefficient to obtain a noise-reduced signal; the noise scene-noise reduction coefficient table is a pre-set table of corresponding noise reduction coefficients that can achieve the optimal noise reduction effect for different noise scenes; the reconstruction module reconstructs the signal according to the noise-reduced signal and the clean signal to obtain the target signal; the recognition module performs speech recognition on the target signal to determine the speech recognition result; and determines the user's order information according to the speech recognition result.

[0151] The beneficial effects of the above technical solution are as follows: The fourth determining module processes the user's voice signal acquired by the second acquisition module into frames and performs Fourier transform processing to obtain a spectrum, which can more clearly and accurately detect the frequency and intensity information of the sound waves, improving the accuracy of the next step in determining whether it is within the preset sound wave range; using its judgment results, the noise intensity characteristics and distribution characteristics of the noisy signal and the clean signal can be obtained more accurately. Using a pre-trained noise recognition model, accurate noise scene recognition of noisy voice signals can be achieved. The noise reduction module uses the method of querying the preset noise scene-noise reduction coefficient table to determine the noise reduction coefficient for the target noise scene to perform noise reduction processing, which can achieve the effect of accurate voice noise reduction and solve the problem of low accuracy in voice recognition.

[0152] According to some embodiments of the present invention, the identification module includes:

[0153] The correction module is used to determine the signal waveform curve corresponding to the target signal, delete the blank segments at both ends of the signal waveform curve included in the signal waveform curve, mark the speech pause position and time, and obtain the corrected signal waveform curve.

[0154] An extraction module is used to extract speech feature sequences from the corrected signal waveform curves included in the corrected signal waveform curve diagram;

[0155] The matching module is used to learn the speech feature sequence based on a preset learning algorithm to obtain phoneme information; and to match the phoneme information with a preset acoustic model to obtain speech recognition results.

[0156] The conversion module is used to convert the speech recognition results into text based on a language model to determine the user's order information.

[0157] The working principle of the above technical solution is as follows: A correction module determines the signal waveform curve corresponding to the target signal, deletes the blank segments at both ends of the signal waveform curve included in the correction waveform curve, and marks the speech pause positions and times to obtain a corrected signal waveform curve; the signal waveform curve is a spectrum diagram of the signal. An extraction module extracts the speech feature sequence from the corrected signal waveform curve included in the corrected signal waveform curve. A matching module learns the speech feature sequence based on a preset learning algorithm to obtain phoneme information; the phoneme information is matched with a preset acoustic model to obtain a speech recognition result. A conversion module converts the speech recognition result into text based on a language model to determine the user's order information.

[0158] The beneficial effects of the above technical solution are: after correcting the target speech, the extraction of speech feature sequences is more conducive to obtaining effective speech sequence features, resulting in more accurate speech recognition results, thereby increasing the user experience.

[0159] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A dispensing drop height adjustment system for a drop-type vending machine, characterized in that, include: The first determining module is used to receive the user's order information, parse the order information, and determine the target product; The first detection module is used to detect the distance information between the target product and the shipping port, including: The acquisition module is used to acquire the distribution information of the storage shelves in the vending machine and generate a shelf distribution map based on the distribution information; The tag module is used for: Determine the storage shelf where the target product is located, and designate it as the target shelf; Mark the target shelves in the shelf distribution map to obtain the marked points; The third determining module is used to determine the distance information between the target product and the shipping outlet based on the location information of the marked point in the shelf distribution map, the location information of the shipping outlet, and the attribute information of the shelf distribution map; Mobile module; Second detection module; The control module, connected to the first detection module, the movement module, and the second detection module respectively, is used for: The system receives distance information sent by the first detection module and determines whether it is greater than a preset distance. When it is determined that the distance information is greater than the preset distance, the system controls the moving module to move to the horizontal plane where the target product will fall. When the moving module moves to the falling horizontal plane and the second detection module determines that the target product has fallen onto the moving module, the moving module is controlled to move to the shipping port; The first determining module includes: The second acquisition module is used to acquire the user's voice signal; The fourth determining module is used for: The speech signal is segmented into frames to obtain sub-speech signals of several fixed length frames; the sub-speech signals are subjected to Fourier transform to obtain frequency domain signals, and the spectrum is determined based on the frequency domain signals. Determine the sound wave frequency in the spectrum diagram and determine whether it is within the preset sound wave frequency range; Determine the sound wave intensity in the spectrum and determine whether it is within a preset sound wave intensity range; Spectrum diagrams with sound wave frequencies and intensities outside the preset range are selected as abnormal spectrum diagrams; the sub-speech signals corresponding to the abnormal spectrum diagrams are taken as noisy signals; the noise intensity characteristics and distribution characteristics of each noisy signal are determined. All spectrograms except those with abnormal spectrograms are considered normal spectrograms; the sub-speech signals corresponding to the normal spectrograms are considered clean signals. The fifth determining module is used for: The noise intensity and distribution features are input into a pre-trained noise recognition model to determine the target noise scene; Noise reduction module, used for: According to the target noise scene, a preset noise scene-noise reduction coefficient table is queried to determine the corresponding noise reduction coefficient. The noisy signal is then denoised according to the noise reduction coefficient to obtain a denoised signal. The reconstruction module is used to: reconstruct the signal based on the noise-reduced signal and the clean signal to obtain the target signal; The recognition module is used for: Perform speech recognition on the target signal and determine the speech recognition result; The user's order information is determined based on the speech recognition results.

2. The dispensing drop height adjustment system of the drop-type vending machine as described in claim 1, characterized in that, The first determining module includes: The first receiving module is used to receive the user's order information; The parsing module is used for: Keyword extraction is performed on the order information to determine target keywords; The target keyword is matched with preset product names in the database to determine the preset product name with the highest matching degree with the target keyword, and the target product is determined based on the preset product name with the highest matching degree with the target keyword.

3. The dispensing drop height adjustment system of the drop-type vending machine as described in claim 2, characterized in that, The parsing module includes: The conversion module is used to convert the order information to obtain the first text; The preprocessing module is used to identify the type of characters included in the first text and remove characters that are consistent with the type included in the blacklist to obtain the second text; The second determining module is used for: Extract the data from the second text to obtain a dataset; Based on the SVM classifier, the text information included in the dataset is classified to obtain several different types of statements; The sentence is segmented using a word segmenter constructed from a dictionary, and the attribute information of each word in the sentence is determined. Based on dependency syntax rules and the attribute information, a diagram illustrating the relationships between various words is determined; The first filtering module is used for: The association diagram is analyzed to determine the connection relationship between each word and other words. The word with the most connection relationships is selected as the target word. The target words are converted into corresponding word vectors. Calculate the distance between the word vector and the word vector of the preset standard word, and select the target words corresponding to the word vectors with a distance less than the preset distance as target keywords.

4. The dispensing drop height adjustment system of the drop-type vending machine as described in claim 1, characterized in that, The mobile module includes: The second receiving module is used to receive instructions sent by the control module; Mobile frame; A drive module is used to drive the moving frame to move up or down according to the instructions.

5. The dispensing drop height adjustment system of the drop-type vending machine as described in claim 1, characterized in that, The second detection module includes a pressure sensor.

6. The dispensing drop height adjustment system of the drop-type vending machine as described in claim 4, characterized in that, Also includes: The third detection module is used to detect whether the driver module is faulty. An alarm module is used to issue an alarm prompt when the third detection module determines that the drive module has malfunctioned.

7. The dispensing drop height adjustment system of the drop-type vending machine as described in claim 6, characterized in that, The third detection module includes: The first acquisition module is used to acquire the sound signal generated by the vibration of the drive module; The judgment module is used for: The sound signal is analyzed to obtain sound features; The sound feature is compared with the standard sound feature corresponding to the standard sound signal. If the difference between the sound feature and the standard sound feature is greater than a preset difference value, it indicates that the driving module has malfunctioned.

8. The dispensing drop height adjustment system of the drop-type vending machine as described in claim 1, characterized in that, The identification module includes: The correction module is used to determine the signal waveform curve corresponding to the target signal, delete the blank segments at both ends of the signal waveform curve included in the signal waveform curve, mark the speech pause position and time, and obtain the corrected signal waveform curve. An extraction module is used to extract speech feature sequences from the corrected signal waveform curves included in the corrected signal waveform curve diagram; The matching module is used to learn the speech feature sequence based on a preset learning algorithm to obtain phoneme information; and to match the phoneme information with a preset acoustic model to obtain speech recognition results. The conversion module is used to convert the speech recognition results into text based on a language model to determine the user's order information.

Images