Instruction processing methods, devices, electronic equipment, vehicles and storage media

Abstract

This disclosure provides an instruction processing method, apparatus, electronic device, vehicle, and storage medium, relating to the field of computer technology, and particularly to the fields of speech recognition, natural language processing, intelligent speech, human-computer interaction, and autonomous driving. The specific implementation scheme includes: acquiring multi-intent voice instructions; obtaining multiple initial single-intent instructions based on the multi-intent voice instructions; detecting and / or adjusting the multiple initial single-intent instructions according to detection rules to obtain multiple target single-intent instructions; and executing each of the multiple target single-intent instructions.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and in particular to the fields of speech recognition, natural language processing, intelligent speech, human-computer interaction, and autonomous driving. Specifically, it relates to an instruction processing method, apparatus, electronic device, vehicle, and storage medium. Background Technology

[0002] With the widespread use of intelligent voice assistants, users are becoming increasingly accustomed to interacting with electronic devices through voice input. However, users often give combined intent commands containing multiple intent instructions in a single voice input, such as "turn on the air conditioner and turn up the temperature." But in existing technologies, electronic devices frequently encounter execution errors and report errors when executing combined intent commands, thereby reducing the user experience. Summary of the Invention

[0003] This disclosure provides an instruction processing method, apparatus, electronic device, vehicle, and storage medium.

[0004] According to a first aspect of this disclosure, an instruction processing method is provided, comprising:

[0005] Acquire multi-intent voice commands;

[0006] Based on multi-intent voice commands, multiple initial single-intent commands are obtained;

[0007] According to the detection rules, multiple initial single intent commands are detected and / or adjusted to obtain multiple target single intent commands.

[0008] Execute each of the multiple target single intent instructions.

[0009] According to a second aspect of this disclosure, an instruction processing apparatus is provided, comprising:

[0010] The first instruction acquisition unit is used to acquire multi-intent voice instructions;

[0011] The second instruction acquisition unit is used to obtain multiple initial single-intent instructions based on multi-intent voice instructions;

[0012] The instruction processing unit is used to detect and / or adjust multiple initial single-intent instructions according to detection rules to obtain multiple target single-intent instructions;

[0013] The instruction execution unit is used to execute each of the multiple target single intent instructions.

[0014] According to a third aspect of this disclosure, an electronic device is provided, comprising:

[0015] At least one processor;

[0016] The memory that is communicatively connected to the at least one processor;

[0017] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method provided in the first aspect of this disclosure.

[0018] According to a fourth aspect of this disclosure, a vehicle is provided, including the electronic equipment provided in a third aspect of this disclosure.

[0019] According to a fifth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are used to cause the computer to perform the method provided according to a first aspect of this disclosure.

[0020] According to a sixth aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the method provided according to a first aspect of this disclosure.

[0021] Using this disclosure can avoid frequent execution errors and error reports when electronic devices execute multi-intent voice commands, thereby enhancing the user experience.

[0022] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0023] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0024] Figure 1 A flowchart illustrating an instruction processing method provided in an embodiment of this disclosure;

[0025] Figure 2 A supplementary illustration of the complete flow of an instruction processing method provided in the embodiments of this disclosure;

[0026] Figure 3 A schematic diagram of a scenario for an instruction processing method provided in an embodiment of this disclosure;

[0027] Figure 4 A schematic structural block diagram of an instruction processing device provided in an embodiment of this disclosure;

[0028] Figure 5 This is a schematic structural block diagram of an electronic device provided in an embodiment of the present disclosure. Detailed Implementation

[0029] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0030] As described in the background section, with the widespread application of intelligent voice assistants, users are increasingly accustomed to interacting with electronic devices (e.g., in-vehicle computers, smart home devices, intelligent voice devices, or other similar computing devices) via voice input. However, users often give combined intent commands containing multiple intent instructions in a single voice input, such as "turn on the air conditioner and turn up the temperature." But in the prior art, electronic devices frequently encounter execution errors and report errors when executing combined intent commands, thereby reducing the user experience.

[0031] Specifically, the inventors have discovered that in the prior art, the reason why electronic devices frequently encounter execution errors and report errors when executing combined intent instructions is that combined intent instructions are raw instructions input by the user, lacking professionalism and rigor. Therefore, combined intent instructions often have execution problems. These execution problems can include at least one of the following: priority error problems, resource conflict problems, logical conflict problems, and execution obstacle problems.

[0032] For example, the combined intent instruction "Start the engine and close the door" includes two intent instructions: "Start the engine" and "Close the door." The "Start the engine" intent instruction comes first, followed by the "Close the door" intent instruction. Therefore, when executing this combined intent instruction, the "Start the engine" intent instruction is executed first, followed by the "Close the door" intent instruction. However, considering safety issues, the "Close the door" intent instruction should be executed first, followed by the "Start the engine" intent instruction. Based on this, it can be understood that the combined intent instruction "Start the engine and close the door" has a priority error. Therefore, when the electronic device executes this combined intent instruction, an execution error and an error message will occur.

[0033] For example, the combined intent instruction "play music and call Zhang San" includes two intent instructions: "play music" and "call Zhang San." These two intent instructions involve the same resource (specifically, an audio resource). Therefore, it is understandable that this combined intent instruction has a resource conflict issue. Consequently, when the electronic device executes this combined intent instruction, an execution error and a report will occur.

[0034] For example, the combined intent instruction "Turn off the air conditioner, turn on the air conditioner" includes two intent instructions: "Turn off the air conditioner" and "Turn on the air conditioner". The intent instruction "Turn off the air conditioner" comes first, and the intent instruction "Turn on the air conditioner" comes second. Therefore, when executing this combined intent instruction, the intent instruction "Turn off the air conditioner" will be executed first, and then the intent instruction "Turn on the air conditioner" will be executed. However, these two intent instructions have a logical conflict. That is, the combined intent instruction "Turn off the air conditioner, turn on the air conditioner" has a logical conflict problem. Therefore, when the electronic device executes this combined intent instruction, an execution error will occur and an error message will be displayed.

[0035] For example, the combined intent instruction "Turn on the air conditioner and increase the fan speed" includes two intent instructions: "Turn on the air conditioner" and "Increase the fan speed". However, the intent instruction "Increase the fan speed" lacks an operand and cannot be executed successfully. In other words, the combined intent instruction "Turn on the air conditioner and increase the fan speed" has an execution obstacle problem. Therefore, when the electronic device executes this combined intent instruction, an execution error will occur and an error message will be displayed.

[0036] To address the above problems, this disclosure provides an instruction processing method that can be applied to electronic devices. The electronic device can be a terminal device. Here, the terminal device can be an in-vehicle computer, a smart home device, a smart voice device, or other similar computing devices. The following will be combined with... Figure 1 The flowchart shown illustrates an instruction processing method provided by an embodiment of this disclosure. It should be noted that although a logical order is shown in the flowchart, in some cases, the steps shown or described in the flowchart may be executed in a different order.

[0037] Step S101: Obtain multi-intent voice commands.

[0038] Multi-intent voice commands are combined intent commands, which can include multiple intent commands given by the user in a single voice input, and each of the multiple intent commands can be in voice form. For example, the multi-intent voice command "Start the engine and close the car door" includes the voice intent command "Start the engine" and the voice intent command "Close the car door". Another example is the multi-intent voice command "Play music and call Zhang San", which includes the voice intent command "Play music" and the voice intent command "Call Zhang San".

[0039] Step S102: Based on multi-intent voice commands, obtain multiple initial single-intent commands.

[0040] In one example, after obtaining a multi-intent voice command, Automatic Speech Recognition (ASR) and Natural Language Understanding (NLU) techniques can be used to process the multi-intent voice command, resulting in multiple initial single-intent commands, each of which is in text form. For example, given the multi-intent voice command "Start the engine and close the car door," processing it with ASR and NLU techniques yields the text-based initial single-intent command "Start the engine" and the text-based initial single-intent command "Close the car door." Similarly, given the multi-intent voice command "Play music and call Zhang San," processing it with ASR and NLU techniques yields the text-based initial single-intent command "Play music" and the text-based initial single-intent command "Call Zhang San."

[0041] Step S103: According to the detection rules, multiple initial single intent commands are detected and / or adjusted to obtain multiple target single intent commands.

[0042] The detection rules are preset and instruct how to detect multiple initial single-intent instructions. If no execution problems exist with the multiple initial single-intent instructions, they are treated as multiple target single-intent instructions; or, if execution problems exist, the multiple initial single-intent instructions are adjusted to obtain multiple target single-intent instructions. Here, execution problems can include at least one of the following: priority errors, resource conflicts, logical conflicts, and execution obstacles.

[0043] Furthermore, in this embodiment of the disclosure, the priority error problem may be that the initial execution order of multiple initial single intent instructions is inconsistent with the priority order determined based on the priority detection rule; the resource conflict problem may be that there are multiple first target instructions involving the same resource among the multiple initial single intent instructions; the logical conflict problem may be that there are multiple second target instructions with logical conflicts among the multiple initial single intent instructions; the execution obstacle problem may be that there is a third target instruction that depends on the specified auto-start item, a fourth target instruction that lacks integrity, or a fifth target instruction that depends on the specified user confirmation item among the multiple initial single intent instructions.

[0044] In this embodiment of the disclosure, when there are execution problems with multiple initial single intent instructions, the multiple initial single intent instructions can be adjusted to solve the execution problems of multiple initial single intent instructions and obtain multiple target single intent instructions.

[0045] Step S104: Execute each of the multiple target single intent instructions.

[0046] After receiving multiple target single-intent instructions, the instruction distribution module can treat each of these instructions as a pending instruction and distribute it to the corresponding vertical business module for execution. These vertical business modules can include window control modules, in-vehicle hardware control modules, media resource control modules, and voice call control modules.

[0047] The instruction processing method provided in this disclosure allows for the acquisition of multi-intent voice instructions and the generation of multiple initial single-intent instructions based on these instructions. Then, according to detection rules, these initial single-intent instructions are detected and / or adjusted to obtain multiple target single-intent instructions. This ensures that the multiple target single-intent instructions do not have execution problems before each of the multiple target single-intent instructions is executed. In this way, when an electronic device executes multi-intent voice instructions, it instead executes each of the multiple target single-intent instructions, and since the multiple target single-intent instructions do not have execution problems, it avoids the frequent occurrence of execution errors and error reports when the electronic device executes multi-intent voice instructions, thereby enhancing the user experience.

[0048] In some optional implementations, step S102, namely, "obtaining multiple initial single-intent commands based on multi-intent voice commands", may include:

[0049] Convert multi-intent voice commands into multi-intent text commands;

[0050] The multi-intent text instruction is parsed to obtain multiple initial single-intent instructions.

[0051] In one example, ASR technology can be used to convert multi-intent speech commands into multi-intent text commands, and NLU technology can be used to parse the multi-intent text commands to obtain multiple initial single-intent commands.

[0052] It should be noted that in the above example, before parsing the multi-intent text instructions to obtain multiple initial single-intent instructions, the multi-intent text instructions can be preprocessed to remove redundant words, such as conjunctions and modal particles.

[0053] For example, given the multi-intent text instruction "Open navigation, turn down the volume, and call Zhang San," after preprocessing this multi-intent text instruction to remove redundant words (i.e., "and" and "ha"), we can obtain the preprocessed multi-intent text instruction "Open navigation, turn down the volume, and call Zhang San." Based on this, it can be understood that in the above example, "parsing the multi-intent text instruction to obtain multiple initial single-intent instructions" can include: parsing the preprocessed multi-intent text instruction to obtain multiple initial single-intent instructions.

[0054] It should also be noted that in the above embodiments, each of the multiple initial single intent instructions may have a specific data structure. For example, each of the multiple initial single intent instructions includes a single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent.

[0055] The vertical information to which a single user intent belongs can be one of the following: window control, in-vehicle hardware control, media resource control, voice call control, etc.

[0056] Here, individual user intents under window controls can include page control, on / off control of various applications, etc.; individual user intents under in-vehicle hardware control can include air conditioning control, lighting control, window control, etc.; individual user intents under media resource control can include video playback control, playback control of audio resources (e.g., music player, radio, etc.); individual user intents under voice telephone control can include making or hanging up a phone call, etc.

[0057] Based on the above, in one example, "parsing multi-intent text instructions to obtain multiple initial single-intent instructions" may include:

[0058] Multi-intent text commands are broken down into multiple single-intent text commands;

[0059] Each single intent text instruction among multiple single intent text instructions is taken as a target text instruction. The target text instruction is parsed to obtain the single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent.

[0060] Based on the single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent, an initial single intent instruction corresponding to the target text instruction is obtained.

[0061] Understandably, in the examples above, the single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent can have specific content or be empty. If any one of the single user intent, the vertical category information to which the single user intent belongs, or the control parameters of the single user intent is empty, it indicates that the initial single intent instruction lacks completeness.

[0062] For example, given the target text instruction "Open navigation", parsing it yields:

[0063] (1) A single user intent, and the single user intent has specific content: GOTO_PAGE (that is, page control);

[0064] (2) The vertical category information to which a single user intent belongs, and the vertical category information to which a single user intent belongs has specific content: windowControl (that is, window control);

[0065] (3) Control parameters for a single user intent, including a first control parameter with specific content: "name": "action", "value": "OPEN", "normalizedValue": "OPEN", and a second control parameter with specific content: "name": "name", "value": "NAVI_SETTING (i.e., navigation page)", "normalizedValue": "NAVI_SETTING".

[0066] Subsequently, based on the above single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent, an initial single intent instruction corresponding to the target text instruction and having a specific data structure can be obtained.

[0067] Through the above implementation methods, in this embodiment of the disclosure, multi-intent voice commands can be converted into multi-intent text commands, and the multi-intent text commands can be parsed to obtain multiple initial single-intent commands. Furthermore, each of the multiple initial single-intent commands has a specific data structure. This reduces the processing difficulty for each of the multiple initial single-intent commands, thereby improving the execution efficiency of the command processing method. Simultaneously, during step S103, the accuracy of the detection results and / or adjustment results for the multiple initial single-intent commands can be ensured, further preventing frequent execution errors and error reports when the electronic device executes multi-intent voice commands, thus enhancing the user experience.

[0068] Furthermore, when parsing multi-intent text commands to obtain multiple initial single-intent commands, the multi-intent text commands are broken down into multiple single-intent text commands. Each of these single-intent text commands is used as a target text command. The target text command is then parsed to obtain the individual user intent, the vertical category information to which the individual user intent belongs, and the control parameters of the individual user intent. Based on the individual user intent, the vertical category information to which the individual user intent belongs, and the control parameters of the individual user intent, the initial single-intent command corresponding to the target text command is obtained. Clearly, the initial single-intent command has a meticulously divided specific data structure. This further ensures the accuracy of the detection results and / or adjustment results for multiple initial single-intent commands during step S103, thereby further avoiding frequent execution errors and error reports when the electronic device executes multi-intent voice commands, thus enhancing the user experience.

[0069] Furthermore, in this embodiment of the disclosure, the detection rules may include at least one of priority detection rules, resource conflict detection rules, logical conflict detection rules, context relationship detection rules, and user confirmation item detection rules, so as to achieve relatively comprehensive detection of multiple initial single intent instructions.

[0070] The rules include: a priority detection rule to detect priority errors in multiple initial single-intent instructions, specifically checking whether the initial execution order of multiple initial single-intent instructions matches the priority order determined by the priority detection rule; a resource conflict detection rule to detect resource conflicts in multiple initial single-intent instructions, specifically checking whether multiple first target instructions involve the same resource; a logical conflict detection rule to detect logical conflicts in multiple initial single-intent instructions, specifically checking whether multiple second target instructions have logical conflicts; a context relationship detection rule to detect execution obstacles in multiple initial single-intent instructions, specifically checking whether a third target instruction has a dependency on a specified auto-start item, and checking whether a fourth target instruction lacks completeness; and a user confirmation item detection rule to detect execution obstacles in multiple initial single-intent instructions, specifically checking whether a fifth target instruction has a dependency on a specified user confirmation item.

[0071] Furthermore, in some optional implementations, step S103, namely, "detecting and / or adjusting multiple initial single-intent instructions according to detection rules to obtain multiple target single-intent instructions," may include:

[0072] Multiple input commands are detected using the current detection rule among multiple detection rules to obtain detection results; where the multiple input commands are obtained based on multiple initial single intent commands;

[0073] Based on the detection results, multiple output instructions are obtained;

[0074] Based on multiple output instructions, multiple target single intent instructions are obtained.

[0075] The multiple detection rules can be any two or more of the following: priority detection rules, resource conflict detection rules, logical conflict detection rules, context relationship detection rules, and user confirmation item detection rules.

[0076] In one example, "using the current detection rule among multiple detection rules to detect multiple input commands and obtain detection results" can include:

[0077] Obtain multiple detection rules;

[0078] The current detection rule is determined sequentially from multiple detection rules according to the detection order;

[0079] Multiple input commands are detected using the current detection rules to obtain the detection results.

[0080] The detection order can be randomly determined or pre-set.

[0081] In a specific example, multiple detection rules include priority detection rules, resource conflict detection rules, logical conflict detection rules, context relationship detection rules, and user confirmation item detection rules. The detection order corresponding to multiple detection rules can be: priority detection rules → resource conflict detection rules → logical conflict detection rules → context relationship detection rules → user confirmation item detection rules.

[0082] Furthermore, it is understood that in this embodiment of the present disclosure, when the current detection rule is the first detection rule among multiple detection rules, the multiple input instructions corresponding to the current detection rule are multiple initial single intent instructions; when the current detection rule is not the first detection rule among multiple detection rules, the multiple input instructions corresponding to the current detection rule are multiple output instructions corresponding to the previous detection rule of the current detection rule. That is, when the current detection rule is not the first detection rule among multiple detection rules, the output of the previous detection rule of the current detection rule will be used as the input of the current detection rule.

[0083] In yet another example, "obtaining multiple output instructions based on the detection results" could include:

[0084] If the detection results indicate that multiple input instructions have execution problems, the multiple input instructions are adjusted according to the adjustment strategy corresponding to the current detection rule to obtain multiple output instructions.

[0085] The adjustment strategies include at least one of the following:

[0086] Reorder multiple input commands;

[0087] Adjust resources for multiple input commands;

[0088] Partial deletion of multiple input commands;

[0089] Expand instructions from multiple input commands;

[0090] Perform instruction replacement on multiple input instructions.

[0091] In a specific example, the adjustment strategy corresponding to the priority detection rule could be: adjusting the order of multiple input instructions; the adjustment strategy corresponding to the resource conflict detection rule could be: adjusting the resources of multiple input instructions (e.g., adjusting the resource allocation); the adjustment strategy corresponding to the logical conflict detection rule could be: partially deleting multiple input instructions; the adjustment strategy corresponding to the context relationship detection rule could be: expanding and / or replacing multiple input instructions; and the adjustment strategy corresponding to the user confirmation item detection rule could be: expanding multiple input instructions.

[0092] If the current detection rule is not the last of multiple detection rules, the multiple output instructions are provided as new multiple input instructions to the next detection rule of the current detection rule, so that the new multiple input instructions can be detected and / or adjusted using the next detection rule of the current detection rule; if the current detection rule is the last of multiple detection rules, the multiple output instructions are provided as multiple target single intent instructions.

[0093] Through the above implementation methods, in this embodiment of the disclosure, the current detection rule among multiple detection rules can be used to detect multiple input commands obtained from multiple initial single-intent commands, obtain detection results, and based on the detection results, obtain multiple output commands, and then based on the multiple output commands, obtain multiple target single-intent commands. In this way, multiple detection rules can be used to perform relatively comprehensive detection on multiple initial single-intent commands, ensuring the accuracy of the detection and adjustment results for multiple initial single-intent commands. This further avoids frequent execution errors and error reports when electronic devices execute multi-intent voice commands, thereby enhancing the user experience.

[0094] Furthermore, in this embodiment, the current detection rule is determined sequentially from multiple detection rules according to the detection order, and multiple input commands are detected using the current detection rule to obtain the detection result. In other words, when multiple initial single-intent commands are detected and / or adjusted using multiple detection rules, the process is serial. This ensures that the detection results are interactive, further guaranteeing the accuracy of the detection and / or adjustment results for multiple initial single-intent commands. This further avoids frequent execution errors and error reports when electronic devices execute multi-intent voice commands, thereby enhancing the user experience.

[0095] Furthermore, in this embodiment, the current detection rule has a corresponding adjustment strategy. When multiple output instructions are obtained based on the detection results, the multiple input instructions are adjusted according to the adjustment strategy corresponding to the current detection rule to obtain multiple output instructions. This improves detection efficiency, reduces user waiting time, and further enhances the user experience.

[0096] In some optional implementations, the multiple detection rules may include priority detection rules, and the current detection rule is the priority detection rule. Based on this, "using the current detection rule among multiple detection rules to detect multiple input instructions and obtain detection results" may include:

[0097] Obtain the predetermined priority order determined based on priority detection rules;

[0098] When the initial execution order of multiple input instructions is inconsistent with the predetermined priority order, a detection result is obtained to characterize the execution problems of multiple input instructions.

[0099] In the predetermined priority order determined by priority detection rules, security-related single-intent instructions are typically set to high priority (first priority), user-defined high-priority single-intent instructions are set to medium priority (second priority), and other single-intent instructions are set to low priority (third priority). Here, security-related single-intent instructions include telephone (including voice and video calls) instructions and navigation instructions, with telephone instructions having a higher priority than navigation instructions.

[0100] For example, consider multiple input commands, with the initial execution order being: "Play music" → "Call Zhang San" → "Turn off the air conditioner" → "Adjust the air conditioner temperature to 22 degrees" → "Turn on the ambient light" → "Adjust to blue" → "Open navigation". Here, "Call Zhang San" and "Open navigation" are single-intent commands related to safety, and in the predetermined priority order determined by priority detection rules, they have the highest priority. Clearly, the initial execution order of the multiple input commands is inconsistent with the priority order determined by the priority detection rules. Therefore, a detection result can be obtained to characterize the execution problem of the multiple input commands. Here, the execution problem of the multiple input commands is specifically a priority error problem.

[0101] Furthermore, in some optional implementations, when a detection result is obtained to characterize the execution problems of multiple input instructions, and the execution problems of the multiple input instructions are specifically priority error problems, "adjusting the multiple input instructions according to the adjustment strategy corresponding to the current detection rule to obtain multiple output instructions" may include:

[0102] According to the adjustment strategy corresponding to the priority detection rule, multiple input instructions are sequentially adjusted based on a predetermined priority order to obtain multiple output instructions.

[0103] Continuing with the example above, following the adjustment strategy corresponding to the priority detection rules, multiple input instructions are sequentially adjusted based on a predetermined priority order. The resulting multiple output instructions could be: "Call Zhang San" → "Open navigation" → "Play music" → "Turn off air conditioning" → "Adjust air conditioning temperature to 22 degrees" → "Turn on ambient light" → "Adjust to blue".

[0104] Through the above implementation methods, in this embodiment of the disclosure, a predetermined priority order determined based on priority detection rules can be obtained. When the initial execution order of multiple input instructions is inconsistent with the predetermined priority order, a detection result characterizing execution problems in the multiple input instructions is obtained. This allows for the adjustment of the multiple input instructions according to an adjustment strategy corresponding to the priority detection rules, based on the predetermined priority order, to obtain multiple output instructions. In this way, it can be ensured that the execution order of the multiple output instructions is consistent with the predetermined priority order determined based on the priority detection rules, thereby ensuring the security and reliability of the instruction processing method provided in this embodiment of the disclosure.

[0105] In some optional implementations, the multiple detection rules may include resource conflict detection rules, and the current detection rule is a resource conflict detection rule. Based on this, "using the current detection rule among multiple detection rules to detect multiple input commands and obtain detection results" may include:

[0106] Identify the resources involved in each of the multiple input instructions;

[0107] By using resource conflict detection rules to determine that multiple input instructions involve multiple first target instructions that share the same resource, a detection result is obtained to characterize the execution problems of multiple input instructions.

[0108] In this context, each input instruction can involve resources such as audio or video resources. Therefore, multiple input instructions involving the same resource can mean that all input instructions involve audio resources; for example, input instructions related to video playback control, input instructions related to audio resource playback control, and input instructions related to voice call control all involve audio resources. Similarly, multiple input instructions involving the same resource can also mean that all input instructions involve video resources; for example, input instructions related to video playback control and input instructions related to video call control both involve video resources.

[0109] For example, there are multiple input instructions: "Call Zhang San" → "Open navigation" → "Play music" → "Turn off air conditioning" → "Adjust air conditioning temperature to 22 degrees" → "Turn on ambient light" → "Adjust to blue". Among these, "Call Zhang San" and "Play music" are multiple first-target instructions involving the same resource (specifically, audio resources). Therefore, detection results can be obtained to characterize the execution problems of these multiple input instructions. Here, the execution problem of these multiple input instructions is specifically a resource conflict problem.

[0110] Furthermore, in some optional implementations, when the detection results used to characterize the execution problems of multiple input instructions are obtained, and the execution problem of each input instruction is specifically a resource conflict problem, it can be determined that the multiple detection rules also include priority detection rules, and "adjusting the multiple input instructions according to the adjustment strategy corresponding to the current detection rule to obtain multiple output instructions" can include:

[0111] Obtain the predetermined priority order determined based on priority detection rules;

[0112] According to the adjustment strategy corresponding to the resource conflict detection rules, and based on a predetermined priority order, resource adjustments are made to multiple first target instructions among multiple input instructions to obtain multiple output instructions.

[0113] The adjustment strategy corresponding to the resource conflict detection rule aims to allocate a large amount or all of the same resource to input instructions with high priority in a predetermined priority order determined by the priority detection rule (e.g., input instructions related to voice call control and input instructions related to video call control). Specifically, a first supplementary intent instruction can be generated to instruct that the same resource be allocated a large amount or all of it to input instructions with high priority in a predetermined priority order determined by the priority detection rule; or, the same resource can be allocated a small amount or not at all to input instructions with low priority in a predetermined priority order determined by the priority detection rule (e.g., input instructions related to video playback control and input instructions related to audio resource playback control). Specifically, a second supplementary intent instruction can be generated to instruct that the same resource be allocated a small amount or not at all to input instructions with low priority in a predetermined priority order determined by the priority detection rule.

[0114] Continuing with the example above, after obtaining the predetermined priority order determined by the priority detection rules, the resources of multiple first target instructions among multiple input instructions are adjusted according to the adjustment strategy corresponding to the resource conflict detection rules, based on the predetermined priority order. The resulting multiple output instructions can be: "Call Zhang San" → "Open navigation" → "Play music" → "Turn down music volume" → "Turn off air conditioning" → "Adjust air conditioning temperature to 22 degrees" → "Turn on ambient light" → "Adjust to blue"; or they can be: "Call Zhang San" → "Open navigation" → "Play music" → "Pause music" → "Turn off air conditioning" → "Adjust air conditioning temperature to 22 degrees" → "Turn on ambient light" → "Adjust to blue".

[0115] Among them, "lower music volume" is a second supplementary intent instruction, used to instruct that a small amount of audio resources be allocated to the input instruction "play music"; "pause music" is a second supplementary intent instruction, used to instruct that no audio resources be allocated to the input instruction "play music".

[0116] Through the above implementation methods, in this embodiment of the disclosure, the resources involved in each of multiple input instructions can be determined. When resource conflict detection rules are used to determine that multiple first target instructions involving the same resources exist among the multiple input instructions, a detection result characterizing the execution problems of the multiple input instructions is obtained. This allows for the acquisition of a predetermined priority order determined based on priority detection rules. Furthermore, according to the adjustment strategy corresponding to the resource conflict detection rules, resources are adjusted for the multiple first target instructions among the multiple input instructions based on the predetermined priority order to obtain multiple output instructions. In this way, a simple resource adjustment strategy can achieve reasonable resource allocation for multiple first target instructions to adapt to the actual needs of users, thereby further enhancing the user experience.

[0117] In some optional implementations, the multiple detection rules may include logical conflict detection rules, and the current detection rule is a logical conflict detection rule. Based on this, "using the current detection rule among multiple detection rules to detect multiple input instructions and obtain detection results" may include:

[0118] Determine the key logical information of each input instruction among multiple input instructions;

[0119] By using logical conflict detection rules to identify multiple second target instructions with key logical information conflicts among multiple input instructions, a detection result is obtained to characterize the execution problems of multiple input instructions.

[0120] Among the multiple input instructions, the key logical information of each input instruction can be the control parameters of the single user intent included in that input instruction.

[0121] In one example, for two input instructions, such as input instruction A and input instruction B, input instruction A and input instruction B can be identified as two second target instructions with critical logical information conflict if the following logical conditions are met:

[0122] (1) The single user intent included in input instruction A is the same as the single user intent included in input instruction B;

[0123] (2) The vertical category information of the single user intent included in input instruction A is the same as that of the single user intent included in input instruction B.

[0124] (3) The control parameters of the single user intent included in input instruction A are different from those of the single user intent included in input instruction B.

[0125] For example, there are multiple input commands: "Call Zhang San" → "Open navigation" → "Play music" → "Turn down music volume" → "Turn off air conditioning" → "Adjust air conditioning temperature to 22 degrees" → "Turn on ambient lighting" → "Adjust to blue". The single user intent included in the input command "Turn off air conditioning" is "Air conditioning control", the vertical category information to which this single user intent (i.e., "Air conditioning control") belongs is in-vehicle hardware control, and the control parameter for this single user intent (i.e., "Air conditioning control") is "Off". Similarly, the single user intent included in the input command "Adjust air conditioning temperature to 22 degrees" is "Air conditioning control", the vertical category information to which this single user intent (i.e., "Air conditioning control") belongs is in-vehicle hardware control, and the control parameter for this single user intent (i.e., "Air conditioning control") is "Adjust temperature to 22 degrees".

[0126] In other words, the single user intent included in the input command "Turn off the air conditioner" is the same as that included in the input command "Adjust the air conditioner temperature to 22 degrees Celsius." The vertical information to which the single user intent included in the input command "Turn off the air conditioner" belongs is also the same as that to the single user intent included in the input command "Adjust the air conditioner temperature to 22 degrees Celsius." However, the control parameters of the single user intent included in the input command "Turn off the air conditioner" are different from those of the single user intent included in the input command "Adjust the air conditioner temperature to 22 degrees Celsius." Therefore, the input commands "Turn off the air conditioner" and "Adjust the air conditioner temperature to 22 degrees Celsius" can be identified as two second target commands with key logical information conflicts, and detection results are obtained to characterize the execution problems of multiple input commands. Here, the execution problems of multiple input commands are specifically logical conflict problems, and are most likely caused by user slips of the tongue.

[0127] Furthermore, in some optional implementations, when a detection result is obtained to characterize the execution problems of multiple input instructions, and the execution problems of the multiple input instructions are specifically logical conflict problems, "adjusting the multiple input instructions according to the adjustment strategy corresponding to the current detection rule to obtain multiple output instructions" may include:

[0128] According to the adjustment strategy corresponding to the logical conflict detection rule, the non-terminal instructions among the multiple second target instructions are identified as instructions to be deleted, so as to delete the instructions to be deleted from the multiple input instructions and obtain multiple output instructions.

[0129] Continuing with the example above, following the adjustment strategy corresponding to the logic conflict detection rule, the non-last instruction among the multiple second target instructions (i.e., the input instruction "turn off the air conditioner") is identified as the instruction to be deleted, so that the instruction to be deleted is removed from the multiple input instructions. The resulting multiple output instructions can be: "Call Zhang San" → "Open navigation" → "Play music" → "Turn down music volume" → "Adjust air conditioner temperature to 22 degrees" → "Turn on ambient light" → "Adjust to blue".

[0130] Through the above implementation methods, in this embodiment of the disclosure, the key logical information of each input instruction among multiple input instructions can be determined. When multiple second target instructions with key logical information conflicts are identified among the multiple input instructions using logical conflict detection rules, a detection result characterizing the execution problems of the multiple input instructions is obtained. Then, according to the adjustment strategy corresponding to the logical conflict detection rules, non-final instructions among the multiple second target instructions are identified as instructions to be deleted, and these instructions are deleted from the multiple input instructions to obtain multiple output instructions. In this way, electronic devices can avoid executing input instructions generated due to user slips of the tongue, etc., so that the execution results obtained from executing multi-intent voice instructions can better adapt to the user's true meaning, thereby further enhancing the user experience.

[0131] In an optional implementation, the multiple detection rules include context relationship detection rules, and the current detection rule is a context relationship detection rule. Based on this, "using the current detection rule among the multiple detection rules to detect multiple input instructions and obtain detection results" may include:

[0132] Specify the startup item;

[0133] By utilizing contextual relationship detection rules to determine if there is a third target instruction among multiple input instructions that has a dependency relationship with a specified auto-start item, a detection result is obtained to characterize the execution problems of multiple input instructions.

[0134] The designated startup items can be items that electronic devices can automatically control to start. Specifically, they can be items with no security threat or low security threat, such as air conditioner startup items, light startup items, video playback startup items, music players, radios, etc.

[0135] Furthermore, in this embodiment of the disclosure, the dependency relationship between a certain input instruction among multiple input instructions and a specified startup item can be understood as the input instruction being executed only when the specified startup item is started.

[0136] For example, there are multiple input commands: "Call Zhang San" → "Open navigation" → "Play music" → "Turn down music volume" → "Adjust air conditioning temperature to 22 degrees" → "Turn on ambient light" → "Change to blue". Among these, "Adjust air conditioning temperature to 22 degrees" is a third target command that depends on a specified startup item (specifically, the air conditioning startup item). Therefore, detection results can be obtained to characterize the execution problems of these multiple input commands. Here, the execution problems of these multiple input commands are specifically execution obstacle problems (which can be understood as the inability to adjust the air conditioning temperature to 22 degrees without turning on the air conditioning).

[0137] Furthermore, in some optional implementations, when a detection result is obtained to characterize the execution problems of multiple input instructions, and the execution problems of the multiple input instructions are specifically execution obstacles caused by the existence of a third target instruction that has a dependency relationship with a specified auto-starting item among the multiple input instructions, "adjusting the multiple input instructions according to the adjustment strategy corresponding to the current detection rule to obtain multiple output instructions" may include:

[0138] Generate startup instructions corresponding to the specified startup item;

[0139] Based on the adjustment strategy corresponding to the context relationship detection rules, multiple input instructions are expanded to obtain multiple output instructions based on the self-starting instruction.

[0140] Among them, the auto-start command corresponding to the specified auto-start item is used to start the auto-start item.

[0141] Continuing the example above, "adjust the air conditioner temperature to 22 degrees" is a third-target instruction that depends on a specified auto-start item (specifically, the air conditioner startup item). Therefore, an auto-start instruction corresponding to the air conditioner startup item can be generated, such as "turn on the air conditioner." Subsequently, based on the auto-start instruction, multiple input instructions are expanded, resulting in multiple output instructions that could be: "call Zhang San" → "open navigation" → "play music" → "lower music volume" → "turn on the air conditioner" → "adjust the air conditioner temperature to 22 degrees" → "turn on the ambient light" → "adjust to blue."

[0142] Through the above implementation methods, in this embodiment of the disclosure, a designated auto-starting item can be determined. When a third target instruction with a dependency relationship to the designated auto-starting item is identified among multiple input instructions using context relationship detection rules, a detection result characterizing the execution problems of the multiple input instructions is obtained. This allows for the generation of an auto-starting instruction corresponding to the designated auto-starting item. Based on the auto-starting instruction, multiple input instructions are expanded according to an adjustment strategy corresponding to the context relationship detection rules to obtain multiple output instructions. Thus, even if a third target instruction with execution obstacles (i.e., an input instruction with a dependency relationship to the designated auto-starting item) exists among multiple input instructions, the third target instruction can be smoothly adjusted to resolve the execution obstacles, thereby ensuring the smooth execution of multi-intent voice instructions and further improving the reliability of the instruction processing method provided in this embodiment of the disclosure.

[0143] In an optional implementation, the multiple detection rules include context relationship detection rules, and the current detection rule is a context relationship detection rule. Based on this, "using the current detection rule among the multiple detection rules to detect multiple input instructions and obtain detection results" may include:

[0144] By utilizing contextual relationship detection rules to determine that a fourth target instruction lacks completeness among multiple input instructions, a detection result is obtained to characterize the execution problems of multiple input instructions.

[0145] For a particular input instruction among multiple input instructions, the lack of completeness may be due to any one of the following being empty: the individual user intent, the vertical category information to which the individual user intent belongs, or the control parameters of the individual user intent.

[0146] For example, consider multiple input instructions: "Call Zhang San" → "Open navigation" → "Play music" → "Turn down music volume" → "Turn on air conditioning" → "Adjust air conditioning temperature to 22 degrees" → "Turn on ambient light" → "Change to blue". The input "Change to blue" lacks a single user intent and the vertical category information to which that intent belongs. Therefore, "Change to blue" can be identified as a fourth target instruction lacking completeness, leading to a detection result characterizing the execution problem of multiple input instructions. Specifically, the execution problem of multiple input instructions is an execution obstacle problem (which can be understood as the input instruction "Change to blue" lacking an operand, thus preventing successful execution).

[0147] Furthermore, in some optional implementations, when a detection result is obtained to characterize the execution problems of multiple input instructions, and the execution problems of the multiple input instructions are specifically execution obstacles caused by the presence of an incomplete fourth target instruction among the multiple input instructions, "adjusting the multiple input instructions according to the adjustment strategy corresponding to the current detection rule to obtain multiple output instructions" may include:

[0148] Supplementing the fourth target instruction yields the complete intent instruction;

[0149] According to the adjustment strategy corresponding to the context relationship detection rule, the fourth target instruction in multiple input instructions is replaced using the complete intent instruction to obtain multiple output instructions.

[0150] In one example, the fourth target instruction can be supplemented by taking into account the context information of the fourth target instruction to obtain a complete intent instruction.

[0151] Continuing with the example above, the fourth target instruction (i.e., the input instruction "adjust to blue") can be supplemented to obtain a complete intent instruction, such as "adjust the ambient light to blue." Then, following the adjustment strategy corresponding to the context detection rules, the complete intent instruction is used to replace the fourth target instruction in multiple input instructions, resulting in multiple output instructions such as: "Call Zhang San" → "Open navigation" → "Play music" → "Turn down music volume" → "Turn on air conditioning" → "Adjust air conditioning temperature to 22 degrees" → "Turn on ambient light" → "Adjust ambient light to blue."

[0152] Through the above implementation methods, in this embodiment of the disclosure, when a fourth target instruction lacking completeness is determined among multiple input instructions using context relationship detection rules, a detection result characterizing the execution problem of multiple input instructions can be obtained. This allows for the supplementation of the fourth target instruction to obtain a complete intent instruction. Furthermore, according to the adjustment strategy corresponding to the context relationship detection rules, the complete intent instruction is used to replace the fourth target instruction among the multiple input instructions, resulting in multiple output instructions. Thus, even if a fourth target instruction with execution obstacles (i.e., an incomplete input instruction) exists among multiple input instructions, the fourth target instruction can be smoothly adjusted to resolve the execution obstacle, thereby ensuring the smooth execution of multi-intent voice instructions and further improving the reliability of the instruction processing method provided in this embodiment of the disclosure.

[0153] In an optional implementation, the multiple detection rules include user confirmation item detection rules, and the current detection rule is the user confirmation item detection rule. Based on this, "using the current detection rule among the multiple detection rules to detect multiple input commands and obtain detection results" may include:

[0154] By utilizing user confirmation item detection rules to determine if there is a fifth target instruction among multiple input instructions that has a dependency relationship with a specified user confirmation item, a detection result is obtained to characterize the execution problems of multiple input instructions.

[0155] Among them, the designated user confirmation item can be an item that the electronic device cannot automatically control to start. Specifically, it can be an item that poses a security threat or has a high security threat, such as a vehicle engine.

[0156] Furthermore, in this embodiment of the disclosure, the dependency relationship between a certain input instruction among multiple input instructions and a specified user confirmation item can be understood as the input instruction being executed only when the specified user confirmation item is activated.

[0157] For example, there are multiple input commands: "Call Zhang San" → "Open navigation" → "Play music" → "Turn down music volume" → "Turn on air conditioning" → "Adjust air conditioning temperature to 22 degrees" → "Turn on ambient lighting" → "Adjust ambient lighting to blue". Among these, "Turn on air conditioning" is the fifth target command that depends on a specified user confirmation item (specifically, the vehicle engine). Therefore, detection results can be obtained to characterize the execution problems of these multiple input commands. Here, the execution problems of these multiple input commands are specifically execution obstacle problems (which can be understood as the inability to turn on the air conditioning without starting the vehicle engine).

[0158] Furthermore, in some optional implementations, when a detection result is obtained to characterize the execution problems of multiple input instructions, and the execution problems of the multiple input instructions are specifically execution obstacles caused by the existence of a fifth target instruction that is dependent on a specified user confirmation item among the multiple input instructions, "adjusting the multiple input instructions according to the adjustment strategy corresponding to the current detection rule to obtain multiple output instructions" may include:

[0159] Generate a prompt message broadcast command corresponding to the specified user confirmation item;

[0160] Based on the adjustment strategy corresponding to the user confirmation item detection rules, and based on the prompt message broadcast instructions, multiple input instructions are expanded to obtain multiple output instructions.

[0161] The prompt message broadcast instruction corresponding to the specified user confirmation item is used to prompt the user to manually activate the specified user confirmation item.

[0162] Continuing the example above, "turn on the air conditioning" is the fifth target instruction that depends on the specified user confirmation item (specifically, the vehicle engine). Therefore, a prompt message broadcast instruction corresponding to the vehicle engine can be generated, such as, "Please confirm whether the vehicle engine is started. If the vehicle engine is not started, please start it manually." Subsequently, based on the prompt message broadcast instruction, multiple input instructions can be expanded to obtain multiple output instructions, such as: "Prompt message broadcast instruction" → "Call Zhang San" → "Open navigation" → "Play music" → "Turn down music volume" → "Turn on the air conditioning" → "Adjust the air conditioning temperature to 22 degrees" → "Turn on the ambient light" → "Adjust the ambient light to blue."

[0163] Through the above implementation methods, in this embodiment of the disclosure, when a fifth target instruction with a dependency relationship to a specified user confirmation item is determined among multiple input instructions using user confirmation item detection rules, a detection result characterizing the execution problem of multiple input instructions can be obtained. This allows for the generation of a prompt message broadcast instruction corresponding to the specified user confirmation item. Furthermore, based on the prompt message broadcast instruction and an adjustment strategy corresponding to the user confirmation item detection rules, the multiple input instructions are expanded to obtain multiple output instructions. Thus, even if a fifth target instruction with an execution obstacle exists among multiple input instructions (i.e., an input instruction with a dependency relationship to a specified user confirmation item), the fifth target instruction can be smoothly adjusted to eliminate the execution obstacle, thereby ensuring the smooth execution of multi-intent voice instructions and further improving the reliability of the instruction processing method provided in this embodiment of the disclosure.

[0164] In an optional implementation, after steps S101, S102, S103, and S104, the instruction processing method may further include:

[0165] After each of the multiple target single intent instructions is successfully executed, an overall feedback message related to the multiple target single intent instructions is generated.

[0166] The overall feedback information is broadcast.

[0167] As described above, in this embodiment of the disclosure, after obtaining multiple target single intent instructions, the instruction distribution module can use each target single intent instruction as an instruction to be executed and distribute the instruction to be executed to the corresponding vertical business module, so that the instruction to be executed can be executed by the vertical business module. The vertical business module can be a window control module, an in-vehicle hardware control module, a media resource control module, or a voice telephone control module.

[0168] In the above implementation, after the instruction distribution module takes each of the multiple target single intent instructions as an instruction to be executed and distributes the instruction to be executed to the corresponding vertical business module, the instruction distribution business module can be used to collect the single intent execution results fed back by each vertical business module, simplify these single intent execution results, generate overall feedback information related to the multiple target single intent instructions, and broadcast the overall feedback information.

[0169] For example, consider multiple single-intention commands: "Turn on navigation" → "Turn on air conditioning" → "Turn on ambient lighting." A typical feedback message might be, "Okay, navigation has been turned on, air conditioning has been turned on, and ambient lighting has been turned on," which is somewhat cumbersome. Using the above implementation method, a holistic feedback message related to multiple single-intention commands can be generated, such as "Navigation, air conditioning, and ambient lighting are all turned on," and this holistic feedback message can be broadcast. In this way, the user is notified through a simplified holistic feedback message related to multiple single-intention commands, further enhancing the user experience.

[0170] Furthermore, it should be noted that in this embodiment of the disclosure, for a single-intent voice command, after the single-intent voice command is successfully executed, a single-intent feedback message can be randomly selected from multiple candidate feedback messages and then broadcast. For example, given the single-intent voice command "Turn on the air conditioner," after the single-intent voice command is successfully executed, a single-intent feedback message can be randomly selected from multiple candidate feedback messages and then broadcast. These multiple candidate feedback messages may include: "Okay, the air conditioner is on for you"; "No problem, the air conditioner is on for you"; "OK, the air conditioner is on for you."

[0171] The following will combine Figure 2 The complete flow of an instruction method provided in the embodiments of this disclosure will be described.

[0172] (1) Obtain multi-intent voice commands.

[0173] Multi-intent voice commands are combined intent commands, which can include multiple intent commands given by the user in a single voice input, and each of these intent commands can be in voice form. For example, the multi-intent voice command "Start the engine and close the car door" includes the voice intent command "Start the engine" and the voice intent command "Close the car door." Another example is the multi-intent voice command "Play music and call Zhang San," which includes the voice intent command "Play music" and the voice intent command "Call Zhang San."

[0174] (2) Based on multi-intent voice commands, multiple initial single-intent commands are obtained.

[0175] For example, a speech processing toolkit (SDK) can be used to process multi-intent speech commands to obtain multiple initial single-intent commands.

[0176] In one example, "deriving multiple initial single-intent commands based on multi-intent voice commands" could include:

[0177] Voice command conversion: Converts multi-intent voice commands into multi-intent text commands;

[0178] Multi-intent text command decomposition: Decompose multi-intent text commands into multiple single-intent text commands;

[0179] Initial Single Intent Command Generation: Each of the multiple single intent text commands is taken as the target text command. The target text command is parsed to obtain the single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent. Based on the single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent, the initial single intent command corresponding to the target text command is obtained.

[0180] (3) Using the instruction detection and adjustment module, multiple initial single intent instructions are detected and / or adjusted according to the detection rules to obtain multiple target single intent instructions.

[0181] The detection rules are preset and instruct how to detect multiple initial single-intent instructions. If no execution problems exist with the multiple initial single-intent instructions, they are treated as multiple target single-intent instructions; or, if execution problems exist, the multiple initial single-intent instructions are adjusted to obtain multiple target single-intent instructions. Here, execution problems can include at least one of the following: priority errors, resource conflicts, logical conflicts, and execution obstacles.

[0182] Furthermore, in this embodiment of the disclosure, the detection rules may include at least one of priority detection rules, resource conflict detection rules, logical conflict detection rules, context relationship detection rules, and user confirmation item detection rules.

[0183] In one example, "detecting and / or adjusting multiple initial single-intent instructions according to detection rules to obtain multiple target single-intent instructions" may include:

[0184] Multiple input commands are detected using the current detection rule among multiple detection rules to obtain detection results; where the multiple input commands are obtained based on multiple initial single intent commands;

[0185] Based on the detection results, multiple output instructions are obtained;

[0186] Based on multiple output instructions, multiple target single intent instructions are obtained.

[0187] The multiple detection rules can be any two or more of the following: priority detection rules, resource conflict detection rules, logical conflict detection rules, context relationship detection rules, and user confirmation item detection rules.

[0188] In a specific example, "using the current detection rule among multiple detection rules to detect multiple input commands and obtain detection results" can include:

[0189] Obtain multiple detection rules;

[0190] The current detection rule is determined sequentially from multiple detection rules according to the detection order;

[0191] Multiple input commands are detected using the current detection rules to obtain the detection results.

[0192] The detection order can be randomly determined or pre-set.

[0193] In a more specific example, the multiple detection rules include priority detection rules, resource conflict detection rules, logical conflict detection rules, context relationship detection rules, and user confirmation item detection rules, and the detection order corresponding to the multiple detection rules can be: priority detection rules → resource conflict detection rules → logical conflict detection rules → context relationship detection rules → user confirmation item detection rules.

[0194] Furthermore, it is understood that in this embodiment of the present disclosure, when the current detection rule is the first detection rule among multiple detection rules, the multiple input instructions corresponding to the current detection rule are multiple initial single intent instructions; when the current detection rule is not the first detection rule among multiple detection rules, the multiple input instructions corresponding to the current detection rule are multiple output instructions corresponding to the previous detection rule of the current detection rule. That is, when the current detection rule is not the first detection rule among multiple detection rules, the output of the previous detection rule of the current detection rule will be used as the input of the current detection rule.

[0195] In yet another specific example, "obtaining multiple output instructions based on the detection results" could include:

[0196] If the detection results indicate that multiple input instructions have execution problems, the multiple input instructions are adjusted according to the adjustment strategy corresponding to the current detection rule to obtain multiple output instructions.

[0197] The adjustment strategies include at least one of the following:

[0198] Reorder multiple input commands;

[0199] Adjust resources for multiple input commands;

[0200] Partial deletion of multiple input commands;

[0201] Expand instructions from multiple input commands;

[0202] Perform instruction replacement on multiple input instructions.

[0203] In a more specific example, the adjustment strategy corresponding to the priority detection rule could be: adjusting the order of multiple input instructions; the adjustment strategy corresponding to the resource conflict detection rule could be: adjusting the resources of multiple input instructions (e.g., adjusting the resource allocation); the adjustment strategy corresponding to the logical conflict detection rule could be: partially deleting multiple input instructions; the adjustment strategy corresponding to the context relationship detection rule could be: expanding and / or replacing multiple input instructions; and the adjustment strategy corresponding to the user confirmation item detection rule could be: expanding multiple input instructions.

[0204] If the current detection rule is not the last of multiple detection rules, the multiple output instructions are provided as new multiple input instructions to the next detection rule of the current detection rule, so that the new multiple input instructions can be detected and / or adjusted using the next detection rule of the current detection rule; if the current detection rule is the last of multiple detection rules, the multiple output instructions are provided as multiple target single intent instructions.

[0205] The following example illustrates the specific process of "detecting and / or adjusting multiple initial single-intent instructions according to the detection rules to obtain multiple target single-intent instructions," using multiple detection rules including priority detection rules, resource conflict detection rules, logical conflict detection rules, context relationship detection rules, and user confirmation item detection rules, and the detection order corresponding to these multiple detection rules being: priority detection rules → resource conflict detection rules → logical conflict detection rules → context relationship detection rules → user confirmation item detection rules.

[0206] The first detection process can be based on priority detection rules to detect and adjust priority errors, specifically including:

[0207] Obtain the predetermined priority order determined based on priority detection rules;

[0208] When the initial execution order of multiple initial single intent instructions is inconsistent with the predetermined priority order, a detection result is obtained to characterize the execution problems of multiple initial single intent instructions. Based on the predetermined priority order, the order of multiple initial single intent instructions is adjusted according to the adjustment strategy corresponding to the priority detection rule to obtain multiple first output instructions.

[0209] Alternatively, if the initial execution order of multiple initial single intent instructions is consistent with the predetermined priority order, a detection result is obtained to characterize that there are no execution problems with the multiple initial single intent instructions, and the multiple initial single intent instructions are used as multiple first output instructions.

[0210] The second detection process can be based on resource conflict detection rules to detect and adjust resource conflict issues, specifically including:

[0211] Determine the resources involved in each of the multiple input instructions (i.e., multiple first output instructions);

[0212] When resource conflict detection rules are used to determine that multiple first target instructions involving the same resources exist among multiple input instructions, a detection result is obtained to characterize the execution problems of multiple input instructions. A predetermined priority order is obtained based on priority detection rules. Then, according to the adjustment strategy corresponding to the resource conflict detection rules, the resources of multiple first target instructions among multiple input instructions are adjusted based on the predetermined priority order to obtain multiple second output instructions.

[0213] Alternatively, by using resource conflict detection rules to determine that there are no multiple first target instructions involving the same resource among multiple input instructions, a detection result is obtained to characterize that there are no execution problems among the multiple input instructions, and the multiple input instructions are treated as multiple second output instructions.

[0214] The third detection process can be based on logic conflict detection rules to detect and adjust for logic conflict issues. Specifically, it can include:

[0215] Determine the key logic information of each input instruction among multiple input instructions (i.e., multiple second output instructions);

[0216] By using logical conflict detection rules to determine that there are multiple second target instructions with key logical information conflicts among multiple input instructions, a detection result is obtained to characterize the execution problem of multiple input instructions. According to the adjustment strategy corresponding to the logical conflict detection rules, the non-terminal instructions among the multiple second target instructions are identified as instructions to be deleted, so as to delete the instructions to be deleted from the multiple input instructions and obtain multiple third output instructions.

[0217] Alternatively, by using logical conflict detection rules to determine that there are no second target instructions with key logical information conflicts among multiple input instructions, a detection result is obtained to characterize that there are no execution problems among multiple input instructions, and the multiple input instructions are treated as multiple third output instructions.

[0218] The fourth detection process can be based on contextual relationship detection rules to detect and adjust for execution obstacles. Specifically, it can include the 4.1 detection process:

[0219] Specify the startup item;

[0220] By using context relationship detection rules to determine that there is a third target instruction that has a dependency relationship with the specified auto-starting item among multiple input instructions (i.e., multiple third output instructions), a detection result is obtained to characterize the execution problem of multiple input instructions, and an auto-starting instruction corresponding to the specified auto-starting item is generated. Then, according to the adjustment strategy corresponding to the context relationship detection rules, multiple input instructions are expanded based on the auto-starting instruction to obtain multiple fourth output instructions.

[0221] Alternatively, by using context relationship detection rules to determine that there is no third target instruction among the multiple input instructions that has a dependency relationship with the specified auto-start item, a detection result is obtained to characterize that there is no execution problem among the multiple input instructions, and the multiple input instructions are treated as multiple fourth output instructions.

[0222] It may also include a 4.2 testing procedure:

[0223] When the context relationship detection rules are used to determine that there is a fourth target instruction lacking completeness among multiple input instructions (i.e., multiple third output instructions), the detection results used to characterize the execution problems of multiple input instructions are obtained, and the fourth target instruction is supplemented to obtain a complete intent instruction. Then, according to the adjustment strategy corresponding to the context relationship detection rules, the fourth target instruction in multiple input instructions is replaced with the complete intent instruction to obtain multiple fourth output instructions.

[0224] Alternatively, by using contextual relationship detection rules to determine that there is no fourth target instruction lacking completeness among multiple input instructions, a detection result is obtained to characterize that there is no execution problem among multiple input instructions, and the multiple input instructions are treated as multiple fourth output instructions.

[0225] It should be noted that the above-mentioned detection process 4.1 and 4.2 can exist independently or simultaneously (e.g., executed serially). That is, when the detection process 4.1 and 4.2 exist simultaneously, multiple third output instructions can be used as inputs to the detection process 4.1, and the output of the detection process 4.1 can be used as inputs to the detection process 4.2, and the output of the detection process 4.2 can be used as multiple fourth output instructions. Alternatively, multiple third output instructions can be used as inputs to the detection process 4.2, and the output of the detection process 4.2 can be used as inputs to the detection process 4.1, and the output of the detection process 4.1 can be used as multiple fourth output instructions. This disclosure does not limit this.

[0226] The fifth detection process can be based on user confirmation item detection rules to detect and adjust for execution obstacles, specifically including:

[0227] By utilizing the user confirmation item detection rules to determine that there is a fifth target instruction among multiple input instructions (i.e., multiple fourth outputs) that has a dependency relationship with the specified user confirmation item, a detection result is obtained to characterize the execution problem of multiple input instructions, and a prompt message broadcast instruction corresponding to the specified user confirmation item is generated. Then, according to the adjustment strategy corresponding to the user confirmation item detection rules, the multiple input instructions are expanded based on the prompt message broadcast instruction to obtain multiple target single intent instructions.

[0228] Alternatively, by using user confirmation item detection rules to determine that there is no fifth target instruction among the multiple input instructions that has a dependency relationship with the specified user confirmation item, a detection result is obtained to characterize that there is no execution problem among the multiple input instructions, and the multiple input instructions are treated as multiple target single intent instructions.

[0229] (4) Execute each of the multiple target single intent instructions.

[0230] After receiving multiple target single-intent instructions, the instruction distribution module can treat each of these instructions as a pending instruction and distribute it to the corresponding vertical business module for execution. These vertical business modules can include window control modules, in-vehicle hardware control modules, media resource control modules, and voice call control modules.

[0231] Subsequently, the instruction distribution business module can be used to collect the single intent execution results fed back by various vertical business modules, and simplify these single intent execution results to generate overall feedback information related to multiple target single intent instructions, and broadcast the overall feedback information.

[0232] Please see Figure 3 This is a schematic diagram of a scenario for an instruction processing method provided in an embodiment of this disclosure.

[0233] As described above, the instruction processing method provided in this disclosure is applied to an electronic device. The electronic device can be a terminal device. Here, the terminal device can be an in-vehicle computer (such as...). Figure 3 As shown, an in-vehicle computer can be installed in a vehicle, a smart home device, a smart voice device, or other similar computing device.

[0234] Electronic devices are used for:

[0235] Acquire multi-intent voice commands;

[0236] Based on multi-intent voice commands, multiple initial single-intent commands are obtained;

[0237] According to the detection rules, multiple initial single intent commands are detected and / or adjusted to obtain multiple target single intent commands.

[0238] Execute each of the multiple target single intent instructions.

[0239] It should be noted that, in the embodiments disclosed herein, Figure 3 The schematic diagrams shown are for illustrative purposes only and are not restrictive. Those skilled in the art can use them as a basis for their own interpretation. Figure 3 The examples may be modified in various obvious ways and / or substitutions, and the resulting technical solutions still fall within the scope of the disclosure of the embodiments of this disclosure.

[0240] To better implement the instruction processing method, this disclosure also provides an instruction processing apparatus that can be integrated into an electronic device. The electronic device can be a terminal device. Here, the terminal device can be an in-vehicle computer, a smart home device, a smart voice device, or other similar computing device. The following will be combined with... Figure 4 The schematic block diagram shown illustrates an instruction processing apparatus 400 provided in a disclosed embodiment.

[0241] Instruction processing device 400, comprising:

[0242] The first instruction acquisition unit 401 is used to acquire multi-intent voice instructions;

[0243] The second instruction acquisition unit 402 is used to obtain multiple initial single-intent instructions based on multi-intent voice instructions;

[0244] The instruction processing unit 403 is used to detect and / or adjust multiple initial single-intent instructions according to detection rules to obtain multiple target single-intent instructions;

[0245] The instruction execution unit 404 is used to execute each of the multiple target single intent instructions.

[0246] In some optional implementations, the detection rules include at least one of priority detection rules, resource conflict detection rules, logical conflict detection rules, context relationship detection rules, and user confirmation item detection rules.

[0247] In some alternative implementations, the instruction processing unit 403 is used for:

[0248] Multiple input commands are detected using the current detection rule among multiple detection rules to obtain detection results; where the multiple input commands are obtained based on multiple initial single intent commands;

[0249] Based on the detection results, multiple output instructions are obtained;

[0250] Based on multiple output instructions, multiple target single intent instructions are obtained.

[0251] In some alternative implementations, the instruction processing unit 403 is used for:

[0252] If the detection results indicate that multiple input instructions have execution problems, the multiple input instructions are adjusted according to the adjustment strategy corresponding to the current detection rule to obtain multiple output instructions.

[0253] In some alternative implementations, the adjustment strategy includes at least one of the following:

[0254] Reorder multiple input commands;

[0255] Adjust resources for multiple input commands;

[0256] Partial deletion of multiple input commands;

[0257] Expand instructions from multiple input commands;

[0258] Perform instruction replacement on multiple input instructions.

[0259] In some alternative implementations, the instruction processing unit 403 is used for:

[0260] Obtain multiple detection rules;

[0261] The current detection rule is determined sequentially from multiple detection rules according to the detection order;

[0262] Multiple input commands are detected using the current detection rules to obtain the detection results.

[0263] In some alternative implementations,

[0264] When the current detection rule is the first of multiple detection rules, the multiple input instructions corresponding to the current detection rule are multiple initial single intent instructions;

[0265] If the current detection rule is not the first detection rule among multiple detection rules, the multiple input instructions corresponding to the current detection rule are the multiple output instructions corresponding to the previous detection rule of the current detection rule.

[0266] In some optional implementations, the multiple detection rules include priority detection rules, and the current detection rule is a priority detection rule; the instruction processing unit 403 is used to:

[0267] Obtain the predetermined priority order determined based on priority detection rules;

[0268] When the initial execution order of multiple input instructions is inconsistent with the predetermined priority order, a detection result is obtained to characterize the execution problems of multiple input instructions.

[0269] In some alternative implementations, the instruction processing unit 403 is used for:

[0270] According to the adjustment strategy corresponding to the priority detection rule, multiple input instructions are sequentially adjusted based on a predetermined priority order to obtain multiple output instructions.

[0271] In some optional implementations, multiple detection rules include resource conflict detection rules, and the current detection rule is a resource conflict detection rule; the instruction processing unit 403 is used to:

[0272] Identify the resources involved in each of the multiple input instructions;

[0273] By using resource conflict detection rules to determine that multiple input instructions involve multiple first target instructions that share the same resource, a detection result is obtained to characterize the execution problems of multiple input instructions.

[0274] In some optional implementations, the multiple detection rules also include priority detection rules; the instruction processing unit 403 is used for:

[0275] Obtain the predetermined priority order determined based on priority detection rules;

[0276] According to the adjustment strategy corresponding to the resource conflict detection rules, and based on a predetermined priority order, resource adjustments are made to multiple first target instructions among multiple input instructions to obtain multiple output instructions.

[0277] In some optional implementations, the multiple detection rules include logical conflict detection rules, and the current detection rule is a logical conflict detection rule; the instruction processing unit 403 is used to:

[0278] Determine the key logical information of each input instruction among multiple input instructions;

[0279] By using logical conflict detection rules to identify multiple second target instructions with key logical information conflicts among multiple input instructions, a detection result is obtained to characterize the execution problems of multiple input instructions.

[0280] In some alternative implementations, the instruction processing unit 403 is used for:

[0281] According to the adjustment strategy corresponding to the logical conflict detection rule, the non-terminal instructions among the multiple second target instructions are identified as instructions to be deleted, so as to delete the instructions to be deleted from the multiple input instructions and obtain multiple output instructions.

[0282] In some optional implementations, multiple detection rules include context relationship detection rules, and the current detection rule is a context relationship detection rule; the instruction processing unit 403 is used to:

[0283] Specify the startup item;

[0284] By utilizing contextual relationship detection rules to determine if there is a third target instruction among multiple input instructions that has a dependency relationship with a specified auto-start item, a detection result is obtained to characterize the execution problems of multiple input instructions.

[0285] In some alternative implementations, the instruction processing unit 403 is used for:

[0286] Generate startup instructions corresponding to the specified startup item;

[0287] Based on the adjustment strategy corresponding to the context relationship detection rules, multiple input instructions are expanded to obtain multiple output instructions based on the self-starting instruction.

[0288] In some optional implementations, multiple detection rules include context relationship detection rules, and the current detection rule is a context relationship detection rule; the instruction processing unit 403 is used to:

[0289] By utilizing contextual relationship detection rules to determine that a fourth target instruction lacks completeness among multiple input instructions, a detection result is obtained to characterize the execution problems of multiple input instructions.

[0290] In some alternative implementations, the instruction processing unit 403 is used for:

[0291] Supplementing the fourth target instruction yields the complete intent instruction;

[0292] According to the adjustment strategy corresponding to the context relationship detection rule, the fourth target instruction in multiple input instructions is replaced using the complete intent instruction to obtain multiple output instructions.

[0293] In some optional implementations, multiple detection rules include user confirmation item detection rules, and the current detection rule is the user confirmation item detection rule; the instruction processing unit 403 is used to:

[0294] By utilizing user confirmation item detection rules to determine if there is a fifth target instruction among multiple input instructions that has a dependency relationship with a specified user confirmation item, a detection result is obtained to characterize the execution problems of multiple input instructions.

[0295] In some alternative implementations, the instruction processing unit 403 is used for:

[0296] Generate a prompt message broadcast command corresponding to the specified user confirmation item;

[0297] Based on the adjustment strategy corresponding to the user confirmation item detection rules, and based on the prompt message broadcast instructions, multiple input instructions are expanded to obtain multiple output instructions.

[0298] In some alternative implementations, the instruction processing unit 403 is used for:

[0299] Convert multi-intent voice commands into multi-intent text commands;

[0300] The multi-intent text instruction is parsed to obtain multiple initial single intent instructions; each of the multiple initial single intent instructions has a specific data structure.

[0301] In some alternative implementations, the instruction processing unit 403 is used for:

[0302] Multi-intent text commands are broken down into multiple single-intent text commands;

[0303] Each single intent text instruction among multiple single intent text instructions is taken as a target text instruction. The target text instruction is parsed to obtain the single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent.

[0304] Based on the single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent, an initial single intent instruction corresponding to the target text instruction is obtained.

[0305] In some optional embodiments, the instruction processing device 400 further includes an information broadcasting unit for:

[0306] After each of the multiple target single intent instructions is successfully executed, an overall feedback message related to the multiple target single intent instructions is generated.

[0307] The overall feedback information is broadcast.

[0308] In this embodiment of the invention, the specific functions and examples of each unit in the instruction processing device 400 can be found in the relevant descriptions of the corresponding steps in the foregoing instruction processing method embodiments, and will not be repeated here.

[0309] The acquisition, storage, and application of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0310] According to embodiments of this disclosure, this disclosure also provides an electronic device, a vehicle, a readable storage medium, and a computer program product.

[0311] Figure 5 A schematic structural block diagram of an example electronic device 500 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as in-vehicle computing devices, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0312] like Figure 5As shown, the electronic device 500 includes a computing unit 501, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. The RAM 503 may also store various programs and data required for the operation of the electronic device 500. The computing unit 501, ROM 502, and RAM 503 are interconnected via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0313] Multiple components in electronic device 500 are connected to I / O interface 505, including: input unit 506, such as keyboard, mouse, etc.; output unit 507, such as various types of renderers, speakers, etc.; storage unit 508, such as disk, optical disk, etc.; and communication unit 509, such as network card, modem, wireless transceiver, etc. Communication unit 509 allows electronic device 500 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0314] The computing unit 501 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. The computing unit 501 performs the various methods and processes described above, such as instruction processing methods. For example, in some embodiments, the instruction processing method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 500 via ROM 502 and / or communication unit 509. When the computer program is loaded into RAM 503 and executed by the computing unit 501, one or more steps of the instruction processing methods described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured as an instruction processing method by any other suitable means (e.g., by means of firmware).

[0315] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transferring data and instructions to the storage system, the at least one input device, and the at least one output device.

[0316] Program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0317] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, ROM, erasable programmable read-only memory (EPROM) or flash memory, optical fibers, compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0318] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a rendering device (e.g., a cathode ray tube (CRT) renderer or a liquid crystal display (LCD)) for rendering information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices are also used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0319] The systems and technologies described herein can be implemented in computing systems that include back-end components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected via digital data communication (e.g., a communication network) of any form or medium. Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0320] A computer system can include client and server components. Clients and servers are generally located far apart and typically interact via a communication network. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, a server in a distributed system, or a server incorporating blockchain technology.

[0321] This disclosure also provides a vehicle, including electronic equipment.

[0322] This disclosure also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to execute an instruction processing method.

[0323] This disclosure also provides a computer program product, including a computer program that implements an instruction processing method when executed by a processor.

[0324] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure is achieved, and this is not limited herein. Furthermore, in this disclosure, relational terms such as "first," "second," and "third" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Additionally, "multiple" in this disclosure can be understood as at least two.

[0325] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. An instruction processing method, comprising: Acquire multi-intent voice commands; Based on the multi-intent voice commands, multiple initial single-intent commands are obtained; According to the detection rules, the multiple initial single intent instructions are detected and / or adjusted to obtain multiple target single intent instructions; wherein, the detection rules include priority detection rules, resource conflict detection rules, logical conflict detection rules, context relationship detection rules, and user confirmation item detection rules; Execute each of the plurality of target single intent instructions; The step of detecting and / or adjusting the multiple initial single-intent instructions according to the detection rules to obtain multiple target single-intent instructions includes: Multiple detection rules are acquired; the current detection rule is determined sequentially from the multiple detection rules according to the detection order; the current detection rule is used to detect multiple input instructions to obtain detection results; wherein, the multiple input instructions are obtained based on multiple initial single intent instructions; If the detection results indicate that the multiple input instructions have execution problems, the multiple input instructions are adjusted according to the adjustment strategy corresponding to the current detection rule to obtain multiple output instructions; wherein, the adjustment strategy includes adjusting the order of the multiple input instructions, adjusting the resources of the multiple input instructions, partially deleting the multiple input instructions, expanding the multiple input instructions, and replacing the multiple input instructions; The multiple target single intent instructions are obtained based on the multiple output instructions.

2. The method according to claim 1, wherein, When the current detection rule is the first detection rule among the plurality of detection rules, the plurality of input instructions corresponding to the current detection rule are the plurality of initial single intent instructions; If the current detection rule is not the first detection rule among the plurality of detection rules, the plurality of input instructions corresponding to the current detection rule are the plurality of output instructions corresponding to the previous detection rule of the current detection rule.

3. The method according to claim 1, wherein, The plurality of detection rules include priority detection rules, and the current detection rule is the priority detection rule; the step of using the current detection rule among the plurality of detection rules to detect multiple input instructions and obtain detection results includes: Obtain the predetermined priority order determined based on the priority detection rules; When the initial execution order of the multiple input instructions is inconsistent with the predetermined priority order, a detection result is obtained to characterize that there is an execution problem with the multiple input instructions.

4. The method according to claim 3, wherein, The adjustment of the multiple input instructions according to the adjustment strategy corresponding to the current detection rule yields multiple output instructions, including: According to the adjustment strategy corresponding to the priority detection rule, the multiple input instructions are sequentially adjusted based on the predetermined priority order to obtain multiple output instructions.

5. The method according to claim 1, wherein, The plurality of detection rules include resource conflict detection rules, and the current detection rule is the resource conflict detection rule; the step of using the current detection rule among the plurality of detection rules to detect multiple input commands and obtain detection results includes: Determine the resources involved in each of the plurality of input instructions; If, by using the resource conflict detection rules, it is determined that there are multiple first target instructions involving the same resource among the multiple input instructions, a detection result is obtained to characterize that there is an execution problem with the multiple input instructions.

6. The method according to claim 5, wherein, The plurality of detection rules also include priority detection rules; the adjustment of the plurality of input instructions according to the adjustment strategy corresponding to the current detection rule to obtain a plurality of output instructions includes: Obtain the predetermined priority order determined based on the priority detection rules; According to the adjustment strategy corresponding to the resource conflict detection rule, and based on the predetermined priority order, the resources of the multiple first target instructions among the multiple input instructions are adjusted to obtain multiple output instructions.

7. The method according to claim 1, wherein, The plurality of detection rules include logical conflict detection rules, and the current detection rule is the logical conflict detection rule; the step of using the current detection rule among the plurality of detection rules to detect multiple input instructions and obtain detection results includes: Determine the key logical information of each of the multiple input instructions; When the logical conflict detection rules are used to determine that there are multiple second target instructions with key logical information conflicts among the multiple input instructions, a detection result is obtained to characterize that there are execution problems in the multiple input instructions.

8. The method according to claim 7, wherein, The adjustment of the multiple input instructions according to the adjustment strategy corresponding to the current detection rule yields multiple output instructions, including: According to the adjustment strategy corresponding to the logical conflict detection rule, the non-end instruction among the plurality of second target instructions is determined as the instruction to be deleted, so as to delete the instruction to be deleted from the plurality of input instructions and obtain a plurality of output instructions.

9. The method according to claim 1, wherein, The plurality of detection rules include context relationship detection rules, and the current detection rule is the context relationship detection rule; the step of using the current detection rule among the plurality of detection rules to detect multiple input instructions and obtain detection results includes: Specify the startup item; By utilizing the context relationship detection rules to determine that there is a third target instruction among the multiple input instructions that has a dependency relationship with the specified auto-start item, a detection result is obtained to characterize that there is an execution problem with the multiple input instructions.

10. The method according to claim 9, wherein, The adjustment of the multiple input instructions according to the adjustment strategy corresponding to the current detection rule yields multiple output instructions, including: Generate a startup instruction corresponding to the specified startup item; According to the adjustment strategy corresponding to the context relationship detection rule, the multiple input instructions are expanded based on the self-starting instruction to obtain multiple output instructions.

11. The method according to claim 1, wherein, The plurality of detection rules include context relationship detection rules, and the current detection rule is the context relationship detection rule; The step of using the current detection rule among the multiple detection rules to detect multiple input commands and obtain detection results includes: By utilizing the aforementioned contextual relationship detection rules to determine that a fourth target instruction lacks completeness among the multiple input instructions, a detection result is obtained to characterize the execution problems of the multiple input instructions.

12. The method according to claim 11, wherein, The adjustment of the multiple input instructions according to the adjustment strategy corresponding to the current detection rule yields multiple output instructions, including: The fourth target instruction is supplemented to obtain a complete intent instruction; According to the adjustment strategy corresponding to the context relationship detection rule, the fourth target instruction in the multiple input instructions is replaced by the complete intent instruction to obtain multiple output instructions.

13. The method according to claim 1, wherein, The plurality of detection rules include user confirmation item detection rules, and the current detection rule is the user confirmation item detection rule; the step of using the current detection rule among the plurality of detection rules to detect multiple input commands and obtain detection results includes: By utilizing the user confirmation item detection rules, if it is determined that there is a fifth target instruction among the multiple input instructions that has a dependency relationship with the specified user confirmation item, a detection result is obtained to characterize that there is an execution problem with the multiple input instructions.

14. The method according to claim 13, wherein, The adjustment of the multiple input instructions according to the adjustment strategy corresponding to the current detection rule yields multiple output instructions, including: Generate a prompt message broadcast command corresponding to the specified user confirmation item; According to the adjustment strategy corresponding to the user confirmation item detection rule, based on the prompt information broadcast instruction, the multiple input instructions are expanded to obtain multiple output instructions.

15. The method according to any one of claims 1 to 14, wherein, The process of obtaining multiple initial single-intent commands based on the multi-intent voice commands includes: Convert the multi-intent voice commands into multi-intent text commands; The multi-intent text instruction is parsed to obtain multiple initial single-intent instructions; wherein each initial single-intent instruction includes a single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent.

16. The method according to claim 15, wherein, The process of parsing the multi-intent text instructions to obtain the multiple initial single-intent instructions includes: The multi-intent text instruction is decomposed into multiple single-intent text instructions; Each of the multiple single-intent text instructions is taken as a target text instruction, and the target text instruction is parsed to obtain a single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent. Based on the single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent, an initial single intent instruction corresponding to the target text instruction is obtained.

17. The method according to claim 1, further comprising: After each of the plurality of target single intent instructions is successfully executed, overall feedback information related to the plurality of target single intent instructions is generated; The overall feedback information is broadcast.

18. An instruction processing apparatus, comprising: The first instruction acquisition unit is used to acquire multi-intent voice instructions; The second instruction acquisition unit is used to obtain multiple initial single-intent instructions based on the multi-intent voice instructions; The instruction processing unit is configured to detect and / or adjust the plurality of initial single intent instructions according to the detection rules to obtain a plurality of target single intent instructions; wherein, the detection rules include priority detection rules, resource conflict detection rules, logical conflict detection rules, context relationship detection rules, and user confirmation item detection rules; An instruction execution unit is used to execute each of the plurality of target single intent instructions; The instruction processing unit is used for: Multiple detection rules are acquired; the current detection rule is determined sequentially from the multiple detection rules according to the detection order; the current detection rule is used to detect multiple input instructions to obtain detection results; wherein, the multiple input instructions are obtained based on multiple initial single intent instructions; If the detection results indicate that the multiple input instructions have execution problems, the multiple input instructions are adjusted according to the adjustment strategy corresponding to the current detection rule to obtain multiple output instructions; wherein, the adjustment strategy includes adjusting the order of the multiple input instructions, adjusting the resources of the multiple input instructions, partially deleting the multiple input instructions, expanding the multiple input instructions, and replacing the multiple input instructions; The multiple target single intent instructions are obtained based on the multiple output instructions.

19. The apparatus according to claim 18, wherein, When the current detection rule is the first detection rule among the plurality of detection rules, the plurality of input instructions corresponding to the current detection rule are the plurality of initial single intent instructions; If the current detection rule is not the first detection rule among the plurality of detection rules, the plurality of input instructions corresponding to the current detection rule are the plurality of output instructions corresponding to the previous detection rule of the current detection rule.

20. The apparatus according to claim 18, wherein, The plurality of detection rules include priority detection rules, and the current detection rule is the priority detection rule; the instruction processing unit is used for: Obtain the predetermined priority order determined based on the priority detection rules; When the initial execution order of the multiple input instructions is inconsistent with the predetermined priority order, a detection result is obtained to characterize that there is an execution problem with the multiple input instructions.

21. The apparatus according to claim 20, wherein, The instruction processing unit is used for: According to the adjustment strategy corresponding to the priority detection rule, the multiple input instructions are sequentially adjusted based on the predetermined priority order to obtain multiple output instructions.

22. The apparatus according to claim 18, wherein, The plurality of detection rules include resource conflict detection rules, and the current detection rule is the resource conflict detection rule; the instruction processing unit is used for: Determine the resources involved in each of the plurality of input instructions; If, by using the resource conflict detection rules, it is determined that there are multiple first target instructions involving the same resource among the multiple input instructions, a detection result is obtained to characterize that there is an execution problem with the multiple input instructions.

23. The apparatus according to claim 22, wherein, The plurality of detection rules also includes priority detection rules; the instruction processing unit is used for: Obtain the predetermined priority order determined based on the priority detection rules; According to the adjustment strategy corresponding to the resource conflict detection rule, and based on the predetermined priority order, the resources of the multiple first target instructions among the multiple input instructions are adjusted to obtain multiple output instructions.

24. The apparatus according to claim 18, wherein, The plurality of detection rules include logical conflict detection rules, and the current detection rule is the logical conflict detection rule; the instruction processing unit is used for: Determine the key logical information of each of the multiple input instructions; When the logical conflict detection rules are used to determine that there are multiple second target instructions with key logical information conflicts among the multiple input instructions, a detection result is obtained to characterize that there are execution problems in the multiple input instructions.

25. The apparatus according to claim 24, wherein, The instruction processing unit is used for: According to the adjustment strategy corresponding to the logical conflict detection rule, the non-end instruction among the plurality of second target instructions is determined as the instruction to be deleted, so as to delete the instruction to be deleted from the plurality of input instructions and obtain a plurality of output instructions.

26. The apparatus according to claim 18, wherein, The plurality of detection rules include context relationship detection rules, and the current detection rule is the context relationship detection rule; the instruction processing unit is used for: Specify the startup item; By utilizing the context relationship detection rules to determine that there is a third target instruction among the multiple input instructions that has a dependency relationship with the specified auto-start item, a detection result is obtained to characterize that there is an execution problem with the multiple input instructions.

27. The apparatus according to claim 26, wherein, The instruction processing unit is used for: Generate a startup instruction corresponding to the specified startup item; According to the adjustment strategy corresponding to the context relationship detection rule, the multiple input instructions are expanded based on the self-starting instruction to obtain multiple output instructions.

28. The apparatus according to claim 18, wherein, The plurality of detection rules include context relationship detection rules, and the current detection rule is the context relationship detection rule; the instruction processing unit is used for: By utilizing the aforementioned contextual relationship detection rules to determine that a fourth target instruction lacks completeness among the multiple input instructions, a detection result is obtained to characterize the execution problems of the multiple input instructions.

29. The apparatus according to claim 28, wherein, The instruction processing unit is used for: The fourth target instruction is supplemented to obtain a complete intent instruction; According to the adjustment strategy corresponding to the context relationship detection rule, the fourth target instruction in the multiple input instructions is replaced by the complete intent instruction to obtain multiple output instructions.

30. The apparatus according to claim 18, wherein, The plurality of detection rules include user confirmation item detection rules, and the current detection rule is the user confirmation item detection rule; the instruction processing unit is used for: By utilizing the user confirmation item detection rules, if it is determined that there is a fifth target instruction among the multiple input instructions that has a dependency relationship with the specified user confirmation item, a detection result is obtained to characterize that there is an execution problem with the multiple input instructions.

31. The apparatus according to claim 30, wherein, The instruction processing unit is used for: Generate a prompt message broadcast command corresponding to the specified user confirmation item; According to the adjustment strategy corresponding to the user confirmation item detection rule, based on the prompt information broadcast instruction, the multiple input instructions are expanded to obtain multiple output instructions.

32. The apparatus according to any one of claims 18 to 31, wherein, The instruction processing unit is used for: Convert the multi-intent voice commands into multi-intent text commands; The multi-intent text instruction is parsed to obtain multiple initial single-intent instructions; wherein each initial single-intent instruction includes a single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent.

33. The apparatus according to claim 32, wherein, The instruction processing unit is used for: The multi-intent text instruction is decomposed into multiple single-intent text instructions; Each of the multiple single-intent text instructions is taken as a target text instruction, and the target text instruction is parsed to obtain a single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent. Based on the single user intent, the vertical category information to which the single user intent belongs, and the control parameters of the single user intent, an initial single intent instruction corresponding to the target text instruction is obtained.

34. The apparatus according to claim 18, further comprising an information broadcasting unit, used for: After each of the plurality of target single intent instructions is successfully executed, overall feedback information related to the plurality of target single intent instructions is generated; The overall feedback information is broadcast.

35. An electronic device comprising: At least one processor; A memory that is communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 17.

36. A vehicle comprising the electronic device of claim 35.

37. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1 to 17.

38. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 17.

Images