Agricultural product wholesale market voice automatic order generation system based on artificial intelligence

Abstract

The invention discloses an agricultural product wholesale market voice automatic order generation system based on artificial intelligence, and relates to the technical field of order management. Comprising an order template generation unit used for generating an order template; the acquisition unit is used for acquiring voice information; the first extraction unit is used for extracting Mel-frequency cepstral coefficient characteristics and residual Mel-frequency cepstral coefficient characteristics of the voice information; the feature splicing unit is used for splicing the features; the operation unit is used for taking the splicing features as input of a speech recognition model to obtain text information; the second extraction unit is used for extracting product names and product quantities in the text information; the matching unit is used for matching the product names with agricultural product names in an agricultural product database to determine agricultural products; and the adding unit is used for adding the agricultural products and the product quantity to the order template to obtain an agricultural product wholesale order. The system disclosed by the invention can improve the order generation efficiency, improve the user experience and improve the operation efficiency of the market.

Description

Technical Field

[0001] The invention belongs to the technical field of order management, and in particular relates to an artificial intelligence-based voice automated order generation system for a wholesale market of agricultural products. Background Art

[0002] As the core link of agricultural product circulation, the operation efficiency of agricultural product wholesale market is directly related to the circulation speed and cost of agricultural products. In traditional agricultural product wholesale market, the generation of agricultural product orders mainly relies on manual input, which requires manual filling of order information on paper or electronic forms, such as agricultural product name, quantity, delivery address, contact information, etc., and then entering this information into the management system for subsequent inventory management, delivery arrangement and other operations.

[0003] However, the process of manually filling in and entering orders is cumbersome and time-consuming, which greatly affects the operational efficiency of the market. Therefore, how to provide an effective solution to improve the efficiency of order generation has become a difficult problem to be solved in the prior art. Summary of the invention

[0004] The purpose of the present invention is to provide an artificial intelligence-based voice automation order generation system for agricultural product wholesale markets to solve the above-mentioned problems existing in the prior art.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] In a first aspect, the present invention provides an artificial intelligence-based voice automated order generation system for agricultural product wholesale markets, comprising:

[0007] An order template generating unit, used to generate an order template in response to an order creation operation initiated by a user;

[0008] An acquisition unit, used to acquire the user's voice information;

[0009] A first extraction unit, used to extract Mel-frequency cepstral coefficient features and residual Mel-frequency cepstral coefficient features of the speech information;

[0010] A feature concatenation unit, used for concatenating the Mel-frequency cepstral coefficient feature with the residual Mel-frequency cepstral coefficient feature to obtain a concatenated feature;

[0011] A computing unit, used for computing the concatenated features as input of a pre-trained speech recognition model to obtain text information corresponding to the speech information;

[0012] A second extraction unit, configured to extract at least one product name and the quantity of products corresponding to the at least one product name from the text information;

[0013] A matching unit, configured to match the at least one product name with the agricultural product names recorded in the agricultural product database, and determine at least one agricultural product matching the at least one product name;

[0014] an adding unit, configured to add the at least one agricultural product and the product quantity corresponding to the at least one agricultural product to the order template to obtain a wholesale order for agricultural products;

[0015] The speech recognition model is trained by using the sample concatenation features obtained by combining the Mel-frequency cepstral coefficient features of the sample speech information and the residual Mel-frequency cepstral coefficient features as sample input and the sample text information corresponding to the sample speech information as sample output.

[0016] Based on the above disclosed content, the present invention provides an artificial intelligence-based voice automated order generation system for agricultural product wholesale markets. By setting an order template generation unit, an acquisition unit, a first extraction unit, a feature combination unit, a calculation unit, a second extraction unit, a matching unit and an adding unit, the order template generation unit is used to generate an order template in response to an order creation operation initiated by a user; the acquisition unit is used to acquire the user's voice information, the first extraction unit is used to extract the Mel-frequency cepstral coefficient feature and the residual Mel-frequency cepstral coefficient feature of the voice information, the feature splicing unit is used to splice the Mel-frequency cepstral coefficient feature with the residual Mel-frequency cepstral coefficient feature to obtain a spliced ​​feature, and the calculation unit is used to use the spliced ​​feature as an input of a pre-trained speech recognition model to calculate and obtain text information corresponding to the voice information, so that the voice features of two dimensions, such as the Mel-frequency cepstral coefficient feature and the residual Mel-frequency cepstral coefficient feature of the voice information, can be extracted, and the text information corresponding to the voice information can be identified after the voice features of the two dimensions are fused and spliced, thereby improving the recognition accuracy of the voice information. The second extraction unit is used to extract at least one product name in the text information and the number of products corresponding to the at least one product name, the matching unit is used to match the at least one product name with the agricultural product name recorded in the agricultural product database, and determine at least one agricultural product that matches the at least one product name, so that the product name in the identified text information can be matched with the agricultural product name recorded in the agricultural product database, thereby matching the standard agricultural product name, and the adding unit is used to add the at least one agricultural product and the number of products corresponding to the at least one agricultural product to the order template to obtain the agricultural product wholesale order, thereby automatically generating the agricultural product wholesale order, wherein the speech recognition model is trained by using the sample splicing feature obtained by combining the Mel frequency cepstral coefficient feature of the sample speech information and the residual Mel frequency cepstral coefficient feature as the sample input, and the sample text information corresponding to the sample speech information as the sample output. Therefore, the voice automation order generation system for the wholesale market of agricultural products based on artificial intelligence provided by the present invention can accurately identify the text information corresponding to the voice information, and match the standard agricultural product name to add to the order template, so that the wholesale order of agricultural products can be automatically generated according to the user's voice, without manually filling in the order information on paper or electronic forms, thereby improving the order generation efficiency, improving the user experience, and improving the market's operating efficiency.

[0017] In a possible design, the AI-based voice automated order generation system for wholesale agricultural product markets also includes:

[0018] A user information acquisition unit, used to acquire user information;

[0019] The adding unit is also used to add the user information to the order template.

[0020] In one possible design, the user information includes user name, user delivery address and / or user contact information.

[0021] In a possible design, when the first extraction unit is used to extract the Mel-frequency cepstral coefficient feature of the speech information, it is specifically used to:

[0022] Performing frame division and windowing processing on the voice information to obtain processed multi-frame signals;

[0023] Performing fast Fourier transform on the processed multi-frame signals to obtain the frequency spectrum of the processed multi-frame signals;

[0024] The frequency spectrum of the processed multi-frame signal is converted into a Mel frequency spectrum and then a logarithmic operation is performed to obtain the logarithmic spectrum of the processed multi-frame signal;

[0025] The processed logarithmic spectra of the multi-frame signals are subjected to discrete cosine transform to obtain Mel-frequency cepstral coefficient features of the speech information.

[0026] In a possible design, when the first extraction unit is used to extract the residual Mel-frequency cepstral coefficient feature of the speech information, it is specifically used to:

[0027] Performing frame division and windowing processing on the voice information to obtain processed multi-frame signals;

[0028] Performing discrete Fourier transform on the multi-frame signals to obtain frequency domain sampling values ​​of the multi-frame signals;

[0029] Calculating the power spectrum density of the multi-frame signal based on the frequency domain sampling of the multi-frame signal;

[0030] Determining linear predictive coding of the multi-frame signal based on the power spectrum density of the multi-frame signal;

[0031] Subtracting the multi-frame signal from the linear prediction coding of the multi-frame signal to obtain a residual signal of the multi-frame signal;

[0032] The residual signal of the multi-frame signal is converted into a Mel frequency spectrum and then cepstrum transformation is performed to obtain the residual Mel frequency cepstrum coefficient characteristics of the speech information.

[0033] In a possible design, the Mel frequency spectrum of the multi-frame signal is subjected to cepstrum transformation according to the following formula (1):

[0034]

[0035] Where N is the number of Mel filters, S n(k, n) represents the response of the kth residual signal after passing through the nth Mel filter, and j is the coefficient order of the residual Mel-frequency cepstrum coefficient.

[0036] In one possible design, the speech recognition model is a Conformer model.

[0037] In a second aspect, the present invention provides a method for generating voice automated orders in a wholesale market of agricultural products based on artificial intelligence, comprising:

[0038] In response to an order creation operation initiated by a user, an order template is generated;

[0039] Get the user's voice information;

[0040] Extracting Mel-frequency cepstral coefficient features and residual Mel-frequency cepstral coefficient features of the speech information;

[0041] Concatenate the Mel-frequency cepstral coefficient feature with the residual Mel-frequency cepstral coefficient feature to obtain a concatenated feature;

[0042] Using the concatenated features as input to a pre-trained speech recognition model to perform operations to obtain text information corresponding to the speech information;

[0043] Extracting at least one product name and the product quantity corresponding to the at least one product name from the text information;

[0044] Matching the at least one product name with the agricultural product names recorded in the agricultural product database to determine at least one agricultural product that matches the at least one product name;

[0045] Adding the at least one agricultural product and the product quantity corresponding to the at least one agricultural product to the order template to obtain an agricultural product wholesale order;

[0046] The speech recognition model is trained by using the sample concatenation features obtained by combining the Mel-frequency cepstral coefficient features of the sample speech information and the residual Mel-frequency cepstral coefficient features as sample input and the sample text information corresponding to the sample speech information as sample output.

[0047] In a third aspect, the present invention provides an electronic device comprising a memory, a processor and a transceiver which are communicatively connected in sequence, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program to execute the method for generating voice automated orders in a wholesale market of agricultural products based on artificial intelligence as described in the second aspect above.

[0048] In a fourth aspect, the present invention provides a computer-readable storage medium having instructions stored thereon. When the instructions are run on a computer, the method for generating voice-automated orders for wholesale agricultural products markets based on artificial intelligence as described in the second aspect is executed.

[0049] In a fifth aspect, the present invention provides a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute the method for generating voice-automated orders in a wholesale agricultural product market based on artificial intelligence as described in the second aspect.

[0050] Beneficial effects:

[0051] The artificial intelligence-based voice automated order generation system for agricultural product wholesale markets provided by the present invention can accurately identify text information corresponding to voice information, and match standard agricultural product names to add to order templates, thereby automatically generating agricultural product wholesale orders based on user voices without the need to manually fill in order information on paper or electronic forms, thereby improving order generation efficiency, enhancing user experience, and improving market operating efficiency, and facilitating practical application and promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 A schematic block diagram of a voice-automated order generation system for a wholesale agricultural product market based on artificial intelligence provided by an embodiment of the present application;

[0053] Figure 2 A flow chart of a method for generating voice-automated orders for wholesale agricultural products markets based on artificial intelligence provided in an embodiment of the present application;

[0054] Figure 3 A schematic block diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0055] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in combination with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structure of the drawings is only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.

[0056] It should be understood that although the terms first, second, etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another unit. For example, a first unit can be referred to as a second unit, and similarly, a second unit can be referred to as a first unit without departing from the scope of the exemplary embodiments of the present invention.

[0057] It should be understood that the term "and / or" that may appear in this article is merely a description of the association relationship between associated objects, indicating that three relationships may exist. For example, A and / or B can represent three situations: A exists alone, B exists alone, and A and B exist at the same time. The term " / and" that may appear in this article describes another type of association object relationship, indicating that two relationships may exist. For example, A / and B can represent two situations: A exists alone, and A and B exist alone. In addition, the character " / " that may appear in this article generally indicates that the previous and next associated objects are in an "or" relationship.

[0058] like Figure 1 As shown, this embodiment provides an artificial intelligence-based voice automation order generation system for agricultural product wholesale markets, which is used to automatically generate agricultural product wholesale orders according to the user's voice. The artificial intelligence-based voice automation order generation system for agricultural product wholesale markets can be set on electronic devices such as smart phones and tablets. The artificial intelligence-based voice automation order generation system for agricultural product wholesale markets includes an order template generation unit, an acquisition unit, a first extraction unit, a feature splicing unit, a calculation unit, a second extraction unit, a matching unit and an adding unit.

[0059] The order template generating unit is used to generate an order template in response to an order creation operation initiated by a user.

[0060] In an embodiment of the present application, a user may manually initiate an order creation operation or initiate an order creation operation through a voice command, which is not specifically limited in the embodiment of the present application. The order template may be a pre-set formatted template.

[0061] The acquisition unit is used to acquire the user's voice information. Specifically, after the order template is generated, the user can enter his order requirements in the form of voice, and the user's voice information can be acquired through the acquisition unit. For example, the voice information can be the voice content of "I need to buy 1,000 kilograms of apples".

[0062] The first extraction unit is used to extract the Mel-frequency cepstral coefficient features and the residual Mel-frequency cepstral coefficient features of the speech information.

[0063] In one or more embodiments, when the first extraction unit is used to extract the Mel-frequency cepstral coefficient features of the speech information, the speech information can be first framed and windowed to obtain a processed multi-frame signal. Framing is to divide the continuous signal into several small time periods (frames) for analysis. Windowing is to process each frame signal so that the edge signal of the frame gradually decreases to zero, thereby reducing the jump between frames. Then, the processed multi-frame signal is subjected to a Fast Fourier Transform (FFT) to obtain the spectrum of the processed multi-frame signal. Then, the spectrum of the processed multi-frame signal can be converted into a Mel frequency spectrum and then a logarithmic operation is performed to obtain the logarithmic spectrum of the processed multi-frame signal. Finally, the logarithmic spectrum of the processed multi-frame signal can be subjected to a discrete cosine transform to obtain the Mel-Frequency Cepstral Coefficients (MFCC) features of the speech information.

[0064] When the first extraction unit is used to extract the residual Mel frequency cepstral coefficient features of the speech information, the speech information can be first framed and windowed to obtain a processed multi-frame signal. The multi-frame signal can be subjected to a discrete Fourier transform (DFT) to obtain a discrete representation of the multi-frame signal in the frequency domain, that is, a frequency domain sampling value of the multi-frame signal. Then, based on the frequency domain sampling of the multi-frame signal, the power spectrum density of the multi-frame signal can be calculated. The power spectrum density of the signal can be expressed as p i =|s i | 2 , where p i represents the power spectrum density of the i-th frame signal, s i Represents the frequency domain sampling of the i-th frame signal. Then, based on the power spectrum density of the multi-frame signal, the linear predictive coding of the multi-frame signal can be determined. Specifically, the power spectrum density of each frame signal can be first subjected to linear predictive coding (Linear Predictive Coding, LPC) analysis to obtain the LPC coefficient, and each frame of the audio signal can be subjected to linear predictive coding using the LPC coefficient to obtain the linear predictive coding of the multi-frame signal. Then, the multi-frame signal can be subtracted from the linear predictive coding of the multi-frame signal to obtain the residual signal of the multi-frame signal. Finally, the residual signal of the multi-frame signal is converted into a Mel-frequency spectrum and then subjected to a cepstrum transform to obtain the residual Mel-frequency cepstrum coefficient feature of the speech information.

[0065] In one or more embodiments, the Mel frequency spectrum of the multi-frame signal may be cepstrum transformed according to the following formula (1):

[0066]

[0067] Where N is the number of Mel filters, S n (k, n) represents the response of the kth residual signal after passing through the nth Mel filter, and j is the coefficient order of the residual Mel-frequency cepstrum coefficient.

[0068] The feature concatenation unit is used to concatenate the Mel-frequency cepstral coefficient feature with the residual Mel-frequency cepstral coefficient feature to obtain a concatenated feature.

[0069] The operation unit is used to use the concatenated features as the input of the pre-trained speech recognition model to perform operations to obtain text information corresponding to the speech information.

[0070] In the embodiment of the present application, a speech recognition model for identifying question information corresponding to speech information is pre-trained, and the speech recognition model can be trained by combining the sample concatenation features obtained by combining the Mel-frequency cepstral coefficient features of the sample speech information and the residual Mel-frequency cepstral coefficient features as sample input, and the sample text information corresponding to the sample speech information as sample output. The speech recognition model can be, but is not limited to, a Conformer model, a Transformer model, etc. In the embodiment of the present application, the speech recognition model is a Conformer model.

[0071] The second extraction unit is used to extract at least one product name and the product quantity corresponding to the at least one product name from the text information.

[0072] The matching unit is used to match at least one product name with the agricultural product name recorded in the agricultural product database, and determine at least one agricultural product that matches the at least one product name. In the embodiment of the present application, an agricultural product database is pre-established, and the agricultural product database records the names of various agricultural products sold. The same agricultural product can have multiple names.

[0073] When matching, at least one product name can be matched with all agricultural product names recorded in the agricultural product database to determine at least one agricultural product that matches the at least one product name. For example, potatoes are called potatoes in some places and are called yangya in some places. When "potatoes", "yangya" or "potatoes" appears in the text information, it can be determined that the matched at least one agricultural product includes potatoes.

[0074] The adding unit is used to add at least one agricultural product and the product quantity corresponding to the at least one agricultural product to the order template to obtain a wholesale order for agricultural products. The product quantity corresponding to the at least one agricultural product refers to the product quantity corresponding to the at least one product name matching the at least one agricultural product.

[0075] In one or more embodiments, the voice automation order generation system for wholesale agricultural products market based on artificial intelligence may further include a user information acquisition unit, which is used to acquire user information, and the adding unit may further be used to add the user information to the order template. The user information may include, but is not limited to, the user name, the user delivery address, and / or the user contact information. In this way, it is convenient to contact the user and arrange the order delivery according to the personal information in the order.

[0076] In summary, the artificial intelligence-based voice automated order generation system for agricultural product wholesale markets provided by the present invention is configured to generate an order template in response to an order creation operation initiated by a user by setting an order template generation unit, an acquisition unit, a first extraction unit, a feature combination unit, a calculation unit, a second extraction unit, a matching unit and an adding unit. The order template generation unit is used to generate an order template in response to an order creation operation initiated by a user, the acquisition unit is used to acquire the voice information of the user, the first extraction unit is used to extract the Mel-frequency cepstral coefficient feature and the residual Mel-frequency cepstral coefficient feature of the voice information, the feature splicing unit is used to splice the Mel-frequency cepstral coefficient feature with the residual Mel-frequency cepstral coefficient feature to obtain a spliced ​​feature, and the calculation unit is used to use the spliced ​​feature as an input of a pre-trained speech recognition model to calculate the voice. The text information corresponding to the information, the second extraction unit is used to extract at least one product name in the text information and the product quantity corresponding to the at least one product name, the matching unit is used to match the at least one product name with the agricultural product name recorded in the agricultural product database, and determine at least one agricultural product that matches the at least one product name, and the adding unit is used to add the at least one agricultural product and the product quantity corresponding to the at least one agricultural product to the order template to obtain a wholesale order for agricultural products, wherein the speech recognition model is trained by combining the Mel frequency cepstral coefficient feature of the sample speech information and the residual Mel frequency cepstral coefficient feature as the sample input, and the sample text information corresponding to the sample speech information as the sample output. Thus, the text information corresponding to the speech information can be accurately identified, and the standard agricultural product name can be matched and added to the order template, so that the wholesale order for agricultural products can be automatically generated according to the user's voice, without manually filling in the order information on paper or electronic forms, thereby improving the efficiency of order generation, improving the user experience, improving the market's operating efficiency, and facilitating practical application and promotion.

[0077] See also Figure 2 The embodiment of the present application provides a method for generating voice-automated orders in an agricultural product wholesale market based on artificial intelligence. The method for generating voice-automated orders in an agricultural product wholesale market based on artificial intelligence may include but is not limited to the following steps S201-S208.

[0078] Step S201. In response to an order creation operation initiated by a user, an order template is generated.

[0079] Step S202: Obtain the user's voice information.

[0080] Step S203: Extract the Mel-frequency cepstral coefficient features of the speech information and the residual Mel-frequency cepstral coefficient features.

[0081] Step S204: Concatenate the Mel-frequency cepstral coefficient feature with the residual Mel-frequency cepstral coefficient feature to obtain a concatenated feature.

[0082] Step S205: Using the concatenated features as input to a pre-trained speech recognition model to perform operations to obtain text information corresponding to the speech information.

[0083] Step S206: Extract at least one product name and the product quantity corresponding to the at least one product name in the text information.

[0084] Step S207: Match the at least one product name with the agricultural product names recorded in the agricultural product database to determine at least one agricultural product that matches the at least one product name.

[0085] Step S208. Add at least one agricultural product and the product quantity corresponding to the at least one agricultural product to the order template to obtain a wholesale order for agricultural products.

[0086] The speech recognition model is trained by using the sample concatenation features obtained by combining the Mel-frequency cepstral coefficient features of the sample speech information and the residual Mel-frequency cepstral coefficient features as sample input and the sample text information corresponding to the sample speech information as sample output.

[0087] The working process, working details and technical effects of the method for generating voice-automated orders for wholesale agricultural products markets based on artificial intelligence provided in the second aspect of this embodiment can be found in the first aspect of the embodiment and will not be described in detail here.

[0088] like Figure 3 As shown, the third aspect of an embodiment of the present application provides an electronic device, comprising a memory, a processor and a transceiver that are communicatively connected in sequence, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program to execute the artificial intelligence-based voice automation order generation method for agricultural products wholesale markets as described in the second aspect of the embodiment.

[0089] For specific examples, the memory may include but is not limited to random access memory (RAM), read-only memory (ROM), flash memory, first-in-first-out memory (FIFO) and / or first-in-last-out memory (FILO), etc.; the processor may be but is not limited to a microprocessor of the STM32F105 series, an ARM (Advanced RISC-Machines), an X86 or other architecture processor, or a processor with an integrated NPU (neural-network processing units); the transceiver may be but is not limited to a WiFi (Wireless Fidelity) wireless transceiver, a Bluetooth wireless transceiver, a General Packet Radio Service (GPRS) wireless transceiver, a ZigBee protocol (a low-power local area network protocol based on the IEEE802.15.4 standard, ZigBee) wireless transceiver, a 3G transceiver, a 4G transceiver and / or a 5G transceiver, etc.

[0090] The fourth aspect of this embodiment provides a computer-readable storage medium storing instructions for the method for generating voice-automated orders for wholesale agricultural products markets based on artificial intelligence as described in the second aspect of the embodiment, that is, the computer-readable storage medium stores instructions, and when the instructions are run on a computer, the method for generating voice-automated orders for wholesale agricultural products markets based on artificial intelligence as described in the second aspect is executed. The computer-readable storage medium refers to a carrier for storing data, which may include but is not limited to a floppy disk, an optical disk, a hard disk, a flash memory, a USB flash drive, and / or a memory stick, etc., and the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0091] The fifth aspect of this embodiment provides a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute the method for generating voice-automated orders for wholesale agricultural products markets based on artificial intelligence as described in the second aspect of the embodiment, wherein the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0092] It should be understood that certain details are provided in the following description to facilitate a complete understanding of the example embodiments. However, it should be understood by those of ordinary skill in the art that the example embodiments can be implemented without these certain details. For example, the system can be shown in a block diagram to avoid obscuring the example with unnecessary details. In other examples, well-known processes, structures, and techniques may not be shown in unnecessary detail to avoid obscuring the example embodiments.

[0093] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An artificial intelligence-based voice automated order generation system for agricultural product wholesale markets, characterized in that: include: An order template generating unit, used to generate an order template in response to an order creation operation initiated by a user; An acquisition unit, used to acquire the user's voice information; A first extraction unit, used to extract Mel-frequency cepstral coefficient features and residual Mel-frequency cepstral coefficient features of the speech information; A feature concatenation unit, used for concatenating the Mel-frequency cepstral coefficient feature with the residual Mel-frequency cepstral coefficient feature to obtain a concatenated feature; A computing unit, used for computing the concatenated features as input of a pre-trained speech recognition model to obtain text information corresponding to the speech information; A second extraction unit, configured to extract at least one product name and the quantity of products corresponding to the at least one product name from the text information; A matching unit, configured to match the at least one product name with the agricultural product names recorded in the agricultural product database, and determine at least one agricultural product matching the at least one product name; an adding unit, configured to add the at least one agricultural product and the product quantity corresponding to the at least one agricultural product to the order template to obtain a wholesale order for agricultural products; The speech recognition model is trained by using the sample concatenation features obtained by combining the Mel-frequency cepstral coefficient features of the sample speech information and the residual Mel-frequency cepstral coefficient features as sample input and the sample text information corresponding to the sample speech information as sample output.

2. The artificial intelligence-based voice automation order generation system for agricultural product wholesale markets according to claim 1 is characterized in that: Also includes: A user information acquisition unit, used to acquire user information; The adding unit is also used to add the user information to the order template.

3. The artificial intelligence-based voice automation order generation system for agricultural product wholesale markets according to claim 2 is characterized in that: The user information includes user name, user delivery address and / or user contact information.

4. The artificial intelligence-based voice automation order generation system for agricultural product wholesale markets according to claim 1 is characterized in that: When the first extraction unit is used to extract the Mel-frequency cepstral coefficient feature of the speech information, it is specifically used to: Performing frame division and windowing processing on the voice information to obtain processed multi-frame signals; Performing fast Fourier transform on the processed multi-frame signals to obtain the frequency spectrum of the processed multi-frame signals; The frequency spectrum of the processed multi-frame signal is converted into a Mel frequency spectrum and then a logarithmic operation is performed to obtain the logarithmic spectrum of the processed multi-frame signal; The processed logarithmic spectra of the multi-frame signals are subjected to discrete cosine transform to obtain Mel-frequency cepstral coefficient features of the speech information.

5. The artificial intelligence-based voice automated order generation system for agricultural product wholesale markets according to claim 1 is characterized in that: When the first extraction unit is used to extract the residual Mel-frequency cepstral coefficient feature of the speech information, it is specifically used to: Performing frame division and windowing processing on the voice information to obtain processed multi-frame signals; Performing discrete Fourier transform on the multi-frame signals to obtain frequency domain sampling values ​​of the multi-frame signals; Calculating the power spectrum density of the multi-frame signal based on the frequency domain sampling of the multi-frame signal; Determining linear predictive coding of the multi-frame signal based on the power spectrum density of the multi-frame signal; Subtracting the multi-frame signal from the linear prediction coding of the multi-frame signal to obtain a residual signal of the multi-frame signal; The residual signal of the multi-frame signal is converted into a Mel frequency spectrum and then cepstrum transformation is performed to obtain the residual Mel frequency cepstrum coefficient characteristics of the speech information.

6. The artificial intelligence-based voice automated order generation system for agricultural product wholesale markets according to claim 5 is characterized in that: The Mel frequency spectrum of the multi-frame signal is subjected to cepstrum transformation according to the following formula (1): Where N is the number of Mel filters, S n (k, n) represents the response of the kth residual signal after passing through the nth Mel filter, and j is the coefficient order of the residual Mel-frequency cepstrum coefficient.

7. The artificial intelligence-based voice automated order generation system for agricultural product wholesale markets according to claim 1 is characterized in that: The speech recognition model is a Conformer model.

8. A method for generating voice automated orders in a wholesale market of agricultural products based on artificial intelligence, characterized in that: include: In response to an order creation operation initiated by a user, an order template is generated; Get the user's voice information; Extracting Mel-frequency cepstral coefficient features and residual Mel-frequency cepstral coefficient features of the speech information; Concatenate the Mel-frequency cepstral coefficient feature with the residual Mel-frequency cepstral coefficient feature to obtain a concatenated feature; Using the concatenated features as input to a pre-trained speech recognition model to perform operations to obtain text information corresponding to the speech information; Extracting at least one product name and the product quantity corresponding to the at least one product name from the text information; Matching the at least one product name with the agricultural product names recorded in the agricultural product database to determine at least one agricultural product that matches the at least one product name; Adding the at least one agricultural product and the product quantity corresponding to the at least one agricultural product to the order template to obtain an agricultural product wholesale order; The speech recognition model is trained by using the sample concatenation features obtained by combining the Mel-frequency cepstral coefficient features of the sample speech information and the residual Mel-frequency cepstral coefficient features as sample input and the sample text information corresponding to the sample speech information as sample output.

9. An electronic device, characterized in that: It includes a memory, a processor and a transceiver which are communicatively connected in sequence, wherein the memory is used to store computer programs, the transceiver is used to send and receive messages, and the processor is used to read the computer program to execute the method for generating voice automated orders in a wholesale market of agricultural products based on artificial intelligence as described in claim 8.

10. A computer program product comprising a computer program or instructions, characterized in that When executed by a computer, the computer program or the instructions implement the method for generating voice-automated orders in a wholesale agricultural product market based on artificial intelligence as described in claim 8.

Images