Trust evaluation method and system of unmanned system, storage medium and electronic device

Abstract

Embodiments of the present application provide a trust degree evaluation method and system of an unmanned system, a storage medium and an electronic device, and relate to the field of unmanned systems. The method comprises: analyzing a communication message sent by a second unmanned device to obtain an evaluation request, wherein the evaluation request at least includes an action event to be executed by a first unmanned device; determining environment context information corresponding to the action event based on the historical trust degree of the action event, and evaluating the trust degree of the action event according to the environment context information to obtain an evaluation result, wherein the environment context information at least includes internal sensing information monitored by a sensing device built in the first unmanned device, and external interaction information generated when the first unmanned device interacts with the second unmanned device; and responding to the action event in a case where it is determined that the evaluation result is used to indicate that the trust degree of the action event is greater than a preset threshold.

Description

Technical Field

[0001] This application relates to the field of unmanned systems, and more specifically, to a trust assessment method, system, storage medium, and electronic device for unmanned systems. Background Technology

[0002] Currently, unmanned systems (USS) are widely used in environmental reconnaissance, astronomical measurement, aerial photography, and logistics distribution. During mission execution, USS systems need to transmit collected data to ground control systems. For example, the ground control system issues mission commands to USS systems. Simultaneously, multiple USS devices in a collaborative operation formation also need to transmit data to each other to achieve coordinated operation. Because current trust assessment mechanisms are only applicable to static entities with clearly defined network structures, while most USS systems are dynamic, moving, and unfamiliar entities with frequent and complex network topologies, existing trust assessment mechanisms cannot evaluate and detect messages transmitted between USS systems, between USS systems and ground control systems, and between USS devices. Furthermore, when USS systems, USS devices, or ground control systems are subjected to cyberattacks, transmitted data may be illegally tampered with, leading to USS systems executing incorrect commands or ground control systems receiving incorrect messages and failing to complete pre-set tasks, resulting in serious consequences. Therefore, when transmitting messages between unmanned systems, between unmanned systems and ground station control systems, and between unmanned devices, it is necessary to conduct a real-time assessment of the trust level of the messages transmitted between unmanned systems, between unmanned systems and ground station control systems, and between unmanned devices to reduce the possibility of mission failure due to illegal tampering of transmitted data.

[0003] Therefore, for related technologies, there is currently no effective solution to the problem of how to assess the trust level of messages between unmanned systems, between unmanned systems and ground station control systems, and between unmanned devices.

[0004] Therefore, it is necessary to improve the relevant technology to overcome the aforementioned defects. Summary of the Invention

[0005] This application provides a method, system, storage medium, and electronic device for assessing the trust level of unmanned systems, at least addressing the problem of how to assess the trust level of messages between unmanned systems, between unmanned systems and ground station control systems, and between unmanned devices.

[0006] According to an aspect of an embodiment of the present application, a trust degree evaluation method of an unmanned system is provided, applied to a first unmanned device, comprising: parsing a communication message sent by a second unmanned device to obtain an evaluation request, wherein the evaluation request at least includes an action event to be executed by the first unmanned device; determining environment context information corresponding to the action event based on a historical trust degree of the action event, and evaluating a trust degree of the action event according to the environment context information to obtain an evaluation result, wherein the environment context information at least includes internal sensing information monitored by a sensing device built in the first unmanned device, and external interaction information generated when the first unmanned device interacts with the second unmanned device, the external interaction information including external environment information of the first unmanned device and device message of the second unmanned device; and in a case where it is determined that the evaluation result is used to indicate that the trust degree of the action event is greater than a preset threshold of trust degree, responding to the action event.

[0007] In an example embodiment, determining the environment context information corresponding to the action event based on the historical trust degree of the action event comprises: obtaining historical data of the action event in a historical scenario from a database, wherein the historical data at least includes a historical trust degree of the action event, a historical event trust level corresponding to the historical trust degree, and a time factor corresponding to the historical trust degree; and generating the environment context information corresponding to the action event based on the historical event trust level and the time factor.

[0008] In an example embodiment, after obtaining the environment context information corresponding to the action event from the database, the method further comprises: converting an information format of the environment context information into a multiple tuple format according to an information conversion rule to obtain converted environment context information, wherein the multiple tuple format at least includes the following format parameters: an information source parameter, an event type parameter, an event occurrence position, an event trust level, and a time factor; and wherein the evaluating the trust degree of the action event according to the environment context information to obtain the evaluation result comprises: evaluating the trust degree of the action event according to the converted environment context information through an evaluation function to obtain the evaluation result.

[0009] In an example embodiment, the trust degree of the action event is evaluated according to the environmental context information by an evaluation function, to obtain an evaluation result, including: determining a first data set corresponding to the external interaction information, and a second data set corresponding to the internal sensing information; calculating a first information entropy of the first data set, and calculating a second information entropy of the second data set; determining a first trust degree corresponding to the external interaction information based on a first data mean of the first data set and the first information entropy, and determining a second trust degree corresponding to the internal sensing information based on a second data mean of the second data set and the second information entropy, wherein the first data mean represents a mean value obtained by averaging all numerical values contained in the first data set, and the second data mean represents a mean value obtained by averaging all numerical values contained in the second data set; and determining the first trust degree and the second trust degree as the evaluation result.

[0010] In an example embodiment, determining the first trust degree and the second trust degree as the evaluation result includes: determining a first weight parameter corresponding to the first trust degree and a second weight parameter corresponding to the second trust degree, wherein a sum value between the first weight parameter and the second weight parameter is a fixed value, the first weight parameter is used to adjust the trust degree of the action event in the external interaction information, and the second weight parameter is used to adjust the trust degree of the action event in the internal sensing information; and adding a first product between the first trust degree and the first weight parameter and a second product between the second trust degree and the second weight parameter to obtain the evaluation result.

[0011] In an example embodiment, the method further comprises: determining a state space set and an action space set corresponding to the environment context information, wherein the action space set comprises a plurality of preset weight parameters, wherein the plurality of preset weight parameters represent an influence degree of the first trust degree in the evaluation result, and the state space set comprises a plurality of groups of state elements; after selecting a current state element from the plurality of groups of state elements each time, updating the first weight parameter and the second weight parameter in an environment state corresponding to the current state element by: in a case of first selecting the plurality of preset weight parameters, selecting a first weight parameter with a maximum value from the plurality of preset weight parameters, determining a selection action occurring when the first weight parameter is selected, and determining a feedback incentive value set for the selection action, wherein the feedback incentive value is used to represent a trust degree of the selection action, in a case that the feedback incentive value is a positive number, representing that the selection action corresponds to positive feedback, and in a case that the feedback incentive value is a negative number, representing that the selection action corresponds to negative feedback; using an iterative function to iteratively calculate the feedback incentive value to obtain a first incentive value corresponding to the selection action; traversing the action space set, selecting weight parameters from the plurality of preset weight parameters multiple times to obtain a plurality of first incentive values; taking a maximum value in the plurality of first incentive values as a target incentive value, and determining a target weight parameter corresponding to the target incentive value, and using the target weight parameter to update the first weight parameter and the second weight parameter.

[0012] In an example embodiment, each group of state elements comprises at least: a first element for representing a first information entropy of the external interaction information, a second element for representing a second information entropy of the internal sensing information, a third element for representing an information relative quantity of the internal sensing information and the external interaction information, and a fourth element for representing a regularity of a trust value of trust information corresponding to the action event, wherein the first element and the second element are discrete variables, and the trust information is from the internal sensing information and the external interaction information.

[0013] According to an aspect of an embodiment of the present application, a trust degree evaluation system of an unmanned system is provided, comprising: a first unmanned device, a second unmanned device; wherein the trust degree evaluation system further comprises a trust degree evaluation module configured on the first unmanned device, configured to parse a communication message received from the second unmanned device to obtain an evaluation request, wherein the evaluation request at least comprises an action event to be executed by the first unmanned device; determine the environment context information corresponding to the action event based on the historical trust degree of the action event, and evaluate the trust degree of the action event according to the environment context information to obtain an evaluation result, wherein the environment context information at least comprises internal sensing information monitored by a sensing device built-in the first unmanned device, and external interaction information generated when the first unmanned device interacts with the second unmanned device, the external interaction information comprising external environment information of the first unmanned device and device message of the second unmanned device; in a case where it is determined that the evaluation result is used to indicate that the trust degree of the action event is greater than a preset threshold of trust degree, respond to the action event.

[0014] In an example embodiment, the trust degree evaluation module is further configured to obtain historical data of the action event in a historical scenario from a database, wherein the historical data at least comprises a historical trust degree of the action event, a historical event trust level corresponding to the historical trust degree, and a time factor corresponding to the historical trust degree; generate the environment context information corresponding to the action event based on the historical event trust level and the time factor.

[0015] In an example embodiment, the first unmanned device is further configured with an information conversion module, configured to, after obtaining the environment context information corresponding to the action event from the database, convert the information format of the environment context information into a multi-tuple format according to an information conversion rule to obtain converted environment context information, wherein the multi-tuple format at least comprises the following format parameters: information source parameter, event type parameter, event occurrence position, event trust level, and time factor; wherein the evaluation result is obtained by evaluating the trust degree of the action event according to the converted environment context information through an evaluation function.

[0016] In an example embodiment, the trustworthiness evaluation module is further configured to determine a first data set corresponding to the external interaction information and a second data set corresponding to the internal sensing information; calculate a first information entropy of the first data set and a second information entropy of the second data set; determine a first trustworthiness corresponding to the external interaction information based on a first data mean of the first data set and the first information entropy, and determine a second trustworthiness corresponding to the internal sensing information based on a second data mean of the second data set and the second information entropy, wherein the first data mean represents a mean value obtained by averaging values of all elements included in the first data set, and the second data mean represents a mean value obtained by averaging values of all elements included in the second data set; and determine the first trustworthiness and the second trustworthiness as the evaluation result.

[0017] In an example embodiment, the trustworthiness evaluation module is further configured to determine a first weight parameter corresponding to the first trustworthiness and a second weight parameter corresponding to the second trustworthiness, wherein a sum of the first weight parameter and the second weight parameter is a constant value, the first weight parameter is used to adjust the trustworthiness of the action event in the external interaction information, and the second weight parameter is used to adjust the trustworthiness of the action event in the internal sensing information; and add a first product of the first trustworthiness and the first weight parameter and a second product of the second trustworthiness and the second weight parameter to obtain the evaluation result.

[0018] In an example embodiment, the trustworthiness evaluation module is further configured to determine a state space set and an action space set corresponding to the environment context information, wherein the action space set comprises a plurality of preset weight parameters, the plurality of preset weight parameters represent an influence degree of the first trustworthiness in the evaluation result, and the state space set comprises a plurality of groups of state elements; after a current state element is selected from the plurality of groups of state elements each time, the first weight parameter and the second weight parameter are updated in an environment state corresponding to the current state element by: in a case where the plurality of preset weight parameters are selected for the first time, selecting a first weight parameter with a maximum value from the plurality of preset weight parameters, determining a selection action that occurs when the first weight parameter is selected, and determining a feedback incentive value set for the selection action, wherein the feedback incentive value is used to represent a trustworthiness degree of the selection action, in a case where the feedback incentive value is a positive number, the selection action corresponds to positive feedback, and in a case where the feedback incentive value is a negative number, the selection action corresponds to negative feedback; performing iterative calculation on the feedback incentive value using an iterative function to obtain a first incentive value corresponding to the selection action; traversing the action space set, selecting weight parameters from the plurality of preset weight parameters multiple times to obtain a plurality of first incentive values; taking a maximum value in the plurality of first incentive values as a target incentive value, determining a target weight parameter corresponding to the target incentive value, and updating the first weight parameter and the second weight parameter using the target weight parameter.

[0019] According to another aspect of the embodiments of the present application, a computer readable storage medium is provided, and the computer readable storage medium stores a computer program. The computer program is configured to execute the trustworthiness evaluation method of the unmanned system when running.

[0020] According to another aspect of the embodiments of the present application, an electronic device is provided, and the electronic device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the trustworthiness evaluation method of the unmanned system through the computer program.

[0021] According to another aspect of the embodiments of the present application, a computer program product is provided, and the computer program product comprises computer instructions. The computer instructions are executed by a processor to implement the trustworthiness evaluation method of the unmanned system.

[0022] By using the first unmanned device to parse the communication message sent by the second unmanned device, the evaluation request of at least the action event to be executed by the first unmanned device is obtained. The environmental context information corresponding to the action event is determined based on the historical trust degree of the action event, and the trust degree of the action event is evaluated according to the environmental context information to obtain an evaluation result. The environmental context information at least includes internal sensing information monitored by the sensing device built in the first unmanned device, and external interaction information generated when the first unmanned device interacts with the second unmanned device. The external interaction information includes external environmental information of the first unmanned device and device message of the second unmanned device. In the case where the evaluation result is used to indicate that the trust degree of the action event is greater than the preset threshold of the trust degree, the action event is responded. The evaluation request of the action event sent by the second unmanned device is parsed by the first unmanned device. The trust evaluation module configured on the first unmanned system evaluates the trust degree of the action event based on the environmental context information. When the calculated trust value of the action event is greater than the set trust value threshold, the first unmanned device responds to the action event. The problem of how to evaluate the message trust degree between unmanned systems, between unmanned systems and ground station control systems, and between unmanned devices is solved, and the accuracy of real-time message transmission is improved. BRIEF DESCRIPTION OF DRAWINGS

[0023] The accompanying drawings, which are included to provide a further understanding of the present application, constitute a part of the present application and illustrate exemplary embodiments of the present application and its description, which are used to explain the present application and do not constitute improper limitations on the present application. In the drawings:

[0024] Figure 1 is a hardware structure block diagram of a computer terminal of the trust degree evaluation method of the unmanned system according to an embodiment of the present application;

[0025] Figure 2 is a flowchart of the trust degree evaluation method of the unmanned system according to an embodiment of the present application;

[0026] Figure 3 is an architecture schematic diagram of the trust degree evaluation system of the unmanned system according to an embodiment of the present application;

[0027] Figure 4 is a structure schematic block diagram of the trust degree evaluation system of the unmanned system according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] In order to make the personnel in the technical field better understand the scheme of the present application, the technical scheme in the embodiments of the present application will be clearly and completely described below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by the person skilled in the art without creative labor should belong to the protection scope of the present application.

[0029] It should be noted that the terms "first", "second" and the like in the specification and claims of the present application and the above-described drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to the process, method, product or device.

[0030] The trust evaluation system of the unmanned system described above can run on a computer terminal, Figure 1 is a hardware structure block diagram of the computer terminal of the trust evaluation method of the unmanned system according to the embodiments of the present application. As shown in Figure 1 , the computer terminal can include one or more (only one is shown in Figure 1 ) processor 102 (the processor 102 can include but is not limited to a microprocessor (Microprocessor Unit, abbreviated as MPU) or a programmable logic device (Programmable logic device, abbreviated as PLD)) and a memory 104 for storing data, in an exemplary embodiment, the computer terminal can also include a transmission device 106 for communication function and an input and output device 108. Those skilled in the art can understand that Figure 1 The structure shown is only schematic, which does not limit the structure of the computer terminal described above. For example, the computer terminal can also include more or fewer components than those shown in Figure 1 , or have a different configuration with the same function as Figure 1 or more functions than Figure 1 .

[0031] The memory 104 can be used to store computer programs, such as software programs of application software and modules, for example, a computer program corresponding to the trust evaluation method of the unmanned system in the embodiments of the present application. The processor 102 can execute various functional applications and data processing, i.e., implement the above method, by running the computer program stored in the memory 104. The memory 104 can include a high-speed random access memory, and can further include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory 104 can further include a memory remotely arranged with respect to the processor 102, which can be connected to the computer terminal through a network. Examples of the above network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0032] The transmission device 106 is used to receive or send data via a network. Specific examples of the above network can include a wireless network provided by a communication provider of the computer terminal. In one example, the transmission device 106 includes a network adapter (NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet in a wireless manner.

[0033] In the embodiments of the present application, a trust evaluation system of an unmanned system is also provided, which is used to implement the following embodiments and preferred embodiments, and will not be described again. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware, or a combination of software and hardware is also possible and is contemplated.

[0034] Next, the trust evaluation method of the unmanned system will be described in combination with the following embodiments.

[0035] Figure 2 is a flowchart of the trust evaluation method of the unmanned system according to the embodiments of the present application, as shown in Figure 2 The steps of the method include:

[0036] In step S202, the received communication message sent by the second unmanned device is parsed to obtain an evaluation request, wherein the evaluation request at least includes an action event to be executed by the first unmanned device.

[0037] Step S204, based on the historical trust degree of the action event, determine the environmental context information corresponding to the action event, and evaluate the trust degree of the action event according to the environmental context information to obtain an evaluation result, wherein the environmental context information at least includes internal sensing information monitored by a sensing device built in the first unmanned device, and external interaction information generated when the first unmanned device interacts with the second unmanned device, the external interaction information including external environmental information of the first unmanned device and device message of the second unmanned device.

[0038] Step S206, in a case where the evaluation result is used to indicate that the trust degree of the action event is greater than a preset trust degree threshold, respond to the action event.

[0039] It should be noted that in the above embodiment, the action event can be understood as various executable events, such as path selection, survey, detection function opening or closing, height adjustment, etc.

[0040] In this embodiment, it can be understood that the trust degree of the action event needs to be evaluated before responding to the action event, and then it is decided whether to respond to the action event based on the evaluation result. If the evaluation result is used to indicate that the trust degree of the action event is greater than the preset trust degree threshold, it means that the evaluated action event is reliable, and the first unmanned device can respond to the action event. If the evaluation result is used to indicate that the trust degree of the action event is less than the preset trust degree threshold, it means that the evaluated action event is unreliable, at this time the second unmanned device can be instructed to resend the request, or the first unmanned device can evaluate the trust degree of the action event again after a period of time.

[0041] Wherein, before executing the above step S202, the first unmanned device can also receive an evaluation request sent by the second unmanned device through an information transmission channel established between the first unmanned device and the second unmanned device.

[0042] The embodiment of the application obtains an evaluation request including at least an action event to be executed by the first unmanned device by analyzing the communication message sent by the second unmanned device; determines the environmental context information corresponding to the action event based on the historical trust degree of the action event, and evaluates the trust degree of the action event according to the environmental context information to obtain an evaluation result, wherein the environmental context information at least includes internal sensing information monitored by a sensing device built in the first unmanned device, and external interaction information generated when the first unmanned device interacts with the second unmanned device, and the external interaction information includes external environmental information of the first unmanned device and device message of the second unmanned device; in a case where it is determined that the evaluation result indicates that the trust degree of the action event is greater than a preset trust degree threshold, the action event is responded, that is, the first unmanned device analyzes the evaluation request including the action event sent by the second unmanned device, the trust evaluation module configured on the first unmanned system evaluates the trust degree of the action event based on the environmental context information, and when the calculated trust value of the action event is greater than a set trust value threshold, the first unmanned device responds to the action event, thereby solving the problem of how to evaluate the message trust degree between unmanned systems, between unmanned systems and ground station control systems, and between unmanned devices, and improving the accuracy of real-time message transmission.

[0043] In an example embodiment, for the implementation process of determining the environmental context information corresponding to the action event based on the historical trust degree of the action event in the above step S204, the following steps can be included: obtaining historical data of the action event in a historical scenario from a database, wherein the historical data at least includes a historical trust degree of the action event, a historical event trust level corresponding to the historical trust degree, and a time factor corresponding to the historical trust degree; generating the environmental context information corresponding to the action event based on the historical event trust level and the time factor.

[0044] In an example embodiment, after the above step S204 is performed, the environmental context information is further format-converted by the following steps: converting the information format of the environmental context information into a multi-tuple format according to an information conversion rule to obtain converted environmental context information, wherein the multi-tuple format at least includes the following format parameters: an information source parameter, an event type parameter, an event occurrence position, an event trust level, and a time factor. Further, in the embodiment, the trust degree of the action event can be further evaluated according to the converted environmental context information by an evaluation function to obtain an evaluation result.

[0045] It should be noted that in the environmental context information, the internal sensing information has different formats due to the differences in sensor types, and the formats of the internal sensing information and the external interaction information also differ, and therefore the above environmental context information needs to be converted into a multiple tuple format according to an information conversion rule to achieve information normalization processing.

[0046] In the embodiment, the information format can be defined as a five tuple (i.e., the above multiple tuple), G = <id, e, l, v1, t1>, where id represents an information source parameter, for example, using the unique identifier of the above ground station, unmanned device or sensor as the information source parameter. e represents an event type. l represents an event occurrence position. v1 represents an event trust level, for example, the value range of v1 can be set to between 0 and 1, when v1 is the minimum value 0, it represents that the event has not occurred or the event is not trustworthy, and when it is the maximum value 1, it represents that the event has occurred or the event is trustworthy. t1 represents a time factor. The scene E of the event e can also be represented by the time factor t1 and the position factor l, for example, the scene E is represented as i = <t1, l>.

[0047] As shown in Figure 3 The environmental context information collected by the message collection and formatting module can be stored in the database by the message collection and formatting module, so that the required information can be obtained from the database at any time, or the action information and the environmental context information are not stored, but are directly sent to the trust evaluation module. Among them, the external interaction information is all information except the internal sensing information of the unmanned device, including the external environmental information of the first unmanned device and the device message of the second unmanned device, which can be specifically understood as the external environmental information of the entity corresponding to the plurality of unmanned devices, and the device information of the communication interaction between the unmanned devices, for example, all interaction information generated by the ground station control system and the unmanned system when the ground station control system sends a control instruction to the unmanned system and receives the communication information of the unmanned system. Each ground station and each unmanned system has a unique identifier. The above internal sensing information includes information collected by various sensors carried by the unmanned system, such as GPS sensors, inertial measurement units, height sensors, infrared sensors, etc., and each sensor has a unique identifier.

[0048] In one example embodiment, for the process of evaluating the trust level of the action event according to the environmental context information by the evaluation function in step S204, the specific process of obtaining the evaluation result can include the following implementation steps: step S11, determining a first data set corresponding to the external interaction information and a second data set corresponding to the internal sensing information; S12, calculating a first information entropy of the first data set and a second information entropy of the second data set; S13, determining a first trust level corresponding to the external interaction information based on a first data mean value of the first data set and the first information entropy, and determining a second trust level corresponding to the internal sensing information based on a second data mean value of the second data set and the second information entropy, wherein the first data mean value represents a mean value obtained by averaging the numerical values of all elements contained in the first data set, and the second data mean value represents a mean value obtained by averaging the numerical values of all elements contained in the second data set; S14, determining the first trust level and the second trust level as the evaluation result.

[0049] Further, in the above embodiments, for example, (e, c) represents an event scene corresponding to the action event e and the scene variable c.

[0050] G out (e, c) and G in (e, c) respectively represent an external data set (i.e. the first data set described above) and an internal data set (i.e. the second data set described above) related to the event scene (e, c), and the process of generating the environmental context information corresponding to the action event based on the historical event trust level and the time factor can be implemented by the following formula:

[0051] K out (e, c) = { <v i = G out (e, c) i .v, t i = G out (e, c) i .t > | i ∈ [1, n out ]}.

[0052] K in (e, c) = { <v i = G in (e, c) i .v, t i = G in (e, c) i .t > | i ∈ [1, n in ]}.

[0053] Wherein, K out (e, c) and K in(e,c) can collectively represent the above-mentioned environmental context information. K out (e,c) is a set containing G out The tuple of factor v and factor t of each element in (e,c) can be understood as the above-mentioned external interaction information, K in (e,c) represents a set containing G in The tuple of factor v and factor t of each element in (e,c) can be understood as the above-mentioned internal sensing information.

[0054] v i and t i represent the scene data set G out (e,c) in the i-th tuple. Wherein, v i represents the above-mentioned historical event trust level, t i represents the above-mentioned time factor.

[0055] n out and n in represent the number of elements in the set G out (e,c) and G in (e,c), respectively. Further, it can be known that:

[0056]

[0057] Optionally, in an exemplary embodiment, the first trust degree f out (e,c) can be calculated by the following formula: Wherein, i.e. the above-mentioned first data mean, is obtained by averaging all elements v in K out (e,c), H(K out (e,c)) is the above-mentioned first information entropy, i.e. the entropy of all elements v.

[0058] Optionally, in an exemplary embodiment, the second trust degree f in (e,c) can be calculated by the following formula: Wherein, i.e. the above-mentioned second data mean, is obtained by averaging all elements v in K in (e,c), H(K in (e,c)) is the above-mentioned second information entropy, representing the entropy of all elements v.

[0059] Optionally, in the above-mentioned embodiment, since the unmanned system can be unknown to other unmanned systems and ground stations during the execution of the task, it is necessary to calculate the trust degree of the collected information. According to the information entropy principle, the formula in the above-mentioned embodiment is K out (e,c), K inThe more chaotic the value distribution of element v in the calculation formula of f (e, c) is, the smaller the final trust degree is, and the evaluation result is closer to 0.5. On the contrary, the more uniform the distribution is, the larger the final trust degree is, i.e. closer to the average value of element v.

[0060] Optionally, in the above embodiments, The calculation formula of f (e, c) is as follows:

[0061]

[0062] The calculation formula of f (e, c) is as follows:

[0063]

[0064] H(K out (e, c)) is as follows:

[0065]

[0066] H(K in (e, c)) is as follows:

[0067]

[0068] Wherein, n out and n in respectively represent the number of elements in set G out (e, c) and G in (e, c), and p x is the probability of element v in set K out (e, c) taking value x. The more the entropy value H(K out (e, c) approaches 1, the more chaotic the value distribution of v in set K out (e, c) is, and the value of f out (e, c) is closer to 0.5. When the entropy value H(K out (e, c) approaches 0, the more uniform the value distribution of v in set K out (e, c) is, and the value of f out (e, c) is closer to the average value of set v. It should be noted that when the number of elements v in set K out (e, c) is zero, 0.5 is taken as the default value of the first trust degree.

[0069] In an example embodiment, for step S14, the following process can be further implemented: determining a first weight parameter corresponding to the first trust degree and a second weight parameter corresponding to the second trust degree, wherein a sum value between the first weight parameter and the second weight parameter is a constant value, the first weight parameter is used to adjust the trust degree of the action event in the external interaction information, and the second weight parameter is used to adjust the trust degree of the action event in the internal sensing information; adding a first product between the first trust degree and the first weight parameter and a second product between the second trust degree and the second weight parameter to obtain the evaluation result.

[0070] Further, in the present embodiment, the evaluation result V final (e,c) can be determined according to the following formula: wherein (e,c) represents an event scene corresponding to an action event e and a scene variable c, V out (e,c) can be understood as the first trust degree, V in (e,c) can be understood as the second trust degree, which can be understood as the first weight parameter, which can be understood as the second weight parameter.

[0071] Further, as mentioned above, which can be understood as the first weight parameter, which can be understood as the second weight parameter. Since the environment of the unmanned system is in a state of dynamic change, the collected information will also change dynamically. For example, the change of the environment caused by the uncertainty of the object communicated by the unmanned system will affect the accuracy of the internal environment information collected by the sensor, and therefore it is necessary to comprehensively consider the internal and external factors to accurately evaluate the trust degree of the event.

[0072] In addition, it should be noted that The value of V

[0073] In an example embodiment, further provided are other technical solutions, specifically comprising: step S21, determining a state space set and an action space set corresponding to the environment context information, wherein the action space set comprises a plurality of preset weight parameters, wherein the plurality of preset weight parameters represent an influence degree of the first trust degree in the evaluation result, and the state space set comprises a plurality of groups of state elements; step S22, after selecting a current state element from the plurality of groups of state elements each time, updating the first weight parameter and the second weight parameter in an environment state corresponding to the current state element by: in the case of selecting the plurality of preset weight parameters for the first time, selecting a first weight parameter with a maximum value from the plurality of preset weight parameters, determining a selection action occurring when the first weight parameter is selected, and determining a feedback incentive value set for the selection action, wherein the feedback incentive value is used to represent a trust degree of the selection action, in the case of the feedback incentive value being a positive number, representing that the selection action corresponds to positive feedback, and in the case of the feedback incentive value being a negative number, representing that the selection action corresponds to negative feedback; step S23, using an iterative function to iteratively calculate the feedback incentive value to obtain a first incentive value corresponding to the selection action; step S24, traversing the action space set, selecting weight parameters from the plurality of preset weight parameters multiple times to obtain a plurality of first incentive values; step S25, taking a maximum value in the plurality of first incentive values as a target incentive value, determining a target weight parameter corresponding to the target incentive value, and using the target weight parameter to update the first weight parameter and the second weight parameter.

[0074] Wherein the influence degree of the first trust degree in the evaluation result increases with the increase of the preset weight parameter, for example, the plurality of preset weight parameters comprise 0.1, 0.3, 0.5, and the corresponding first trust degree increases in turn, that is, the first trust degree corresponding to 0.5 is the largest.

[0075] In the embodiment, for example, a known action space set AS and a state space set ST (representing a space state with a base number of 22 of all states in the above state space set) are set, an evaluation of feedback after performing an action δ on an event scenario (e, c) in a state s is created Q(s, δ), M(s, δ), s∈ST∩δ(0, 1, 2), wherein Q[s, δ] represents an approximate optimal incentive value (i.e., a feedback incentive value) of performing an action δ when evaluating the trust degree of an event request in a state s, and M[s, δ] represents a feedback incentive value of performing an action δ in a state s. The process of obtaining an optimal action value function V(s, δ) is as follows:

[0076]

[0077] The above step indicates that in the reinforcement learning initialization stage, no internal information is available in the unmanned system, and only external information similar to the instruction can be utilized, and therefore, the Q[s, 2] = 1 is set to initialize the evaluation strategy.

[0078]

[0079] In the formula, the parameter β represents a learning rate in the range of [0, 1], and the greater the value of β, the smaller the influence of the historical evaluation result on the learning process.

[0080]

[0081]

[0082] It can be seen that in the above steps, M(s, δ) records the feedback incentive obtained after performing an action, and V(s, δ) represents the approximate optimization incentive value required for the action. V(s, δ) is iteratively calculated, and the following formula is referred to:

[0083]

[0084] When V(s, δ) is maximum, the weight The maximum element of V(s, δ) can be selected from the action space set (representing the weight value range) {0.2, 0.5, 0.8}, and the optimized weight is calculated. After updating the original weight, dynamic optimization is achieved.

[0085] In an example embodiment, it should be noted that each set of state elements at least includes: a first element for representing a first information entropy of the external interaction information, a second element for representing a second information entropy of the internal sensing information, a third element for representing a relative quantity of information of the internal sensing information and the external interaction information, and a fourth element for representing a regularity of a trust value of trust information corresponding to the action event, wherein the first element and the second element are discrete variables, and the trust information is derived from the internal sensing information and the external interaction information.

[0086] Optionally, in the above embodiment, for example, the action space set is set as AS = {0.2, 0.5, 0.8}, and the unmanned system can only take one value in the action space set as the weight parameter of the value.

[0087] In an example embodiment, for example, the above state space set can be represented as S = <h in ,h out ,N,RG>, wherein h in represents the information entropy of the internal sensing information, hout information entropy of external interaction information, h in and h out is a discrete variable, and the value range is {0, 1}, N represents the information quantity of the internal sensing information and the external interaction information, and RG represents the regularity of the trust degree of the environmental context information corresponding to the action event.

[0088] Optionally, in the above embodiment, h in and h out The value range is set to {0, 1}, 0 represents a low value, and 1 represents a high value. Wherein, the calculation formula of h in The calculation formula of h out represents rounding down.

[0089] Optionally, in the above embodiment, N in the state space set can be calculated by the following calculation formula:

[0090]

[0091] Wherein,

[0092]

[0093] It should be noted that N represents the information quantity of the internal sensing information and the external interaction information, which will affect the trust evaluation of the action event by the unmanned system, so N is set as one of the elements of the state space set. If the unmanned system often executes a specific event in a certain scene, the unmanned system has more information about the scene and the event, that is, the trust degree of the event is higher. If the unmanned system only performs trust evaluation on an event in a certain scene for several times, it means that the trust degree of the event by the unmanned system is lower.

[0094] Optionally, in the above embodiment, the calculation steps of RG in the state space set are as follows:

[0095] Step S1: according to the action event, create an ordered set of corresponding internal sensing information and external interaction information:

[0096] K = K out (e, c) U K in (e, c);

[0097] Wherein, e represents the event type, and c represents the internal environmental information and the external interaction information.

[0098] Step S2: sort according to the time sequence of the factor t of each element in the ordered set K.

[0099] ​​Step S3: Determine the number of elements n = |K| in the ordered set K.

[0100] Step S4: Set the initial value of RG to 1.

[0101] Step S5: Perform the following operation logic:

[0102]

[0103] That is, in the case where n is greater than 1, the value of RG(e, c) can be calculated by the formula If the calculated value of RG(e, c) is greater than 0.5, the value of RG(e, c) is determined to be 0, otherwise the value of RG(e, c) is determined to be 1.

[0104] Optionally, in the above embodiment, when the unmanned system receives an evaluation request including an action event, the environment context information corresponding to the action event can be obtained first, and then the action event is evaluated using the formula and the environment context information.

[0105] In an optional embodiment, as shown in Figure 3 , the trust evaluation system of the unmanned system is applied in the actual scene, and different modules can be specifically set, such as a message collection and formatting module, a trust evaluation module, a strategy optimization module, an unmanned system state control module, etc. However, these modules are collectively used to implement the trust evaluation method of the unmanned system performed by the first unmanned device. The running process of the trust evaluation system of the unmanned system is further explained in combination with the following steps:

[0106] Step 1: The event evaluation request R (i.e. the above evaluation request) is generated by the unmanned system state control module, and the event evaluation request R is sent to the trust evaluation module through the conversion interface. The event evaluation request R can include: event e (i.e. action event), scene E, and the event evaluation request R can also be understood as a request for evaluating the trust degree of the event e in the context of the scene E.

[0107] Step 2: The message collection and formatting module converts the event evaluation request R into a multiple tuple format according to the information conversion rule, and sends it to the trust evaluation module through the conversion interface.

[0108] Step 3: After receiving the event evaluation request R, the trust evaluation module queries the environment context information associated with R from the database.

[0109] Step 4: The trust evaluation module evaluates the action event according to the queried environment context information.

[0110] Step 5: After completing the event evaluation request R, the trust evaluation module sends the evaluation result to the conversion interface for format conversion.

[0111] Step 6: The conversion interface sends the evaluation result to the unmanned system state control module, and the unmanned system state control module adjusts the system parameters of the unmanned system according to the evaluation result to control the unmanned system.

[0112] Step 7: The unmanned system state control module feeds back the evaluation result to the conversion interface.

[0113] Step 8: The conversion interface converts the evaluation result in step 7 into a tuple format and sends it to the policy optimization module, and the reinforcement learning model in the policy optimization module optimizes the evaluation policy (which can be implemented by the reinforcement learning module in the policy optimization module) with the evaluation result as input. Figure 3

[0114] Step 9: The trust evaluation module accesses the policy optimization module to obtain the evaluation policy for evaluating the event evaluation request R.

[0115] Step 10: After the policy optimization module optimizes the evaluation policy according to the feedback of the evaluation result, the calculation function (i.e., the evaluation calculation algorithm) required for evaluation is obtained.

[0116] Step 11: The policy optimization module stores the evaluation result as the environmental context information of the action event in the database for next call.

[0117] Through the above embodiment, the diversity of environmental changes and the uncertainty of the interaction objects of the unmanned equipment are comprehensively considered, the trust degree of the unmanned system information is evaluated in a complex and changing scene, and the reinforcement learning model is introduced for the characteristics of strong maneuverability of the unmanned system and continuous change of the scene, the historical evaluation results can be referred to, so that the evaluation algorithm can be dynamically and real-timely adjusted, and more optimal evaluation results can be output in subsequent similar scenes.

[0118] Optionally, in the optimization process of step 8, the Q-learning reinforcement learning algorithm can also be introduced to learn the evaluation result fed back by the trust evaluation module based on the evaluation request, and the value of the weight parameter can be optimized according to the learning result. Specifically, the learning engine of the policy optimization module shown in Figure 3 can be used to continuously receive the evaluation result fed back from the unmanned system control module, and the evaluation result can also be further used as a reward for the strategy used after the trust evaluation of the event evaluation request R(e, c).

[0119] In this scheme, it is assumed that the trust evaluation result is consistent with the actual demand, so the incentive is set as M = 1, and otherwise the incentive is M = 0.

[0120] ​In the above embodiments, the policy optimization module needs to learn from the evaluation results fed back from the unmanned system control module when performing optimization, therefore, the application introduces environmental context information to reflect the event scenario, the information quality and the information quantity of the event related information, and the application also uses information entropy to represent the information quality. The lower the value of the entropy, the more uniform the information distribution, and the higher the probability of the corresponding event occurrence.

[0121] The method embodiments provided in the embodiments of the application can be executed in the trust degree evaluation system of the unmanned system. In one embodiment, the trust degree evaluation system is combined with the policy optimization module. Figure 4 The trust degree evaluation system of the unmanned system is described. Figure 4 is a structural schematic block diagram of the trust degree evaluation system of the unmanned system according to the embodiments of the application.

[0122] As shown in Figure 4 , the trust degree evaluation system of the unmanned system specifically comprises: a first unmanned device 42 and a second unmanned device 44; wherein the trust degree evaluation system further comprises a trust degree evaluation module configured on the first unmanned device, which is used to parse a communication message sent by the second unmanned device to obtain an evaluation request, wherein the evaluation request at least comprises an action event to be executed by the first unmanned device; environmental context information corresponding to the action event is determined based on the historical trust degree of the action event, and the trust degree of the action event is evaluated according to the environmental context information to obtain an evaluation result, wherein the environmental context information at least comprises internal sensing information monitored by a sensing device built in the first unmanned device, and external interaction information generated when the first unmanned device interacts with the second unmanned device, the external interaction information comprising external environmental information of the first unmanned device and device message of the second unmanned device; in the case where it is determined that the evaluation result is used to indicate that the trust degree of the action event is greater than a preset threshold value of the trust degree, the action event is responded to.

[0123] It should be noted that, in the application, only the first unmanned device is taken as an example to describe the trust degree evaluation method of the unmanned system, and in actual scenarios, any unmanned device can execute the trust degree evaluation method of the unmanned system in the application.

[0124] In one exemplary embodiment, the above-mentioned trust degree evaluation module is further used to obtain historical data of the action event in a historical scenario from a database, wherein the historical data at least comprises a historical trust degree of the action event, a historical event trust level corresponding to the historical trust degree, and a time factor corresponding to the historical trust degree; the environmental context information corresponding to the action event is generated based on the historical event trust level and the time factor.

[0125] In an example embodiment, the first unmanned device is further configured with an information conversion module, configured to convert the information format of the environment context information into a multi-tuple format according to an information conversion rule after obtaining the environment context information corresponding to the action event from the database, to obtain converted environment context information, wherein the multi-tuple format includes at least the following format parameters: information source parameter, event type parameter, event occurrence location, event trust level, time factor; wherein the trust degree of the action event is evaluated according to the converted environment context information to obtain an evaluation result.

[0126] In an example embodiment, the trust degree evaluation module is further configured to determine a first data set corresponding to the external interaction information and a second data set corresponding to the internal sensing information; calculate a first information entropy of the first data set and a second information entropy of the second data set; determine a first trust degree corresponding to the external interaction information based on a first data mean value of the first data set and the first information entropy, and determine a second trust degree corresponding to the internal sensing information based on a second data mean value of the second data set and the second information entropy, wherein the first data mean value represents a mean value obtained by averaging the numerical values of all elements contained in the first data set, and the second data mean value represents a mean value obtained by averaging the numerical values of all elements contained in the second data set; and determine the first trust degree and the second trust degree as the evaluation result.

[0127] In an example embodiment, the trust degree evaluation module is further configured to determine a first weight parameter corresponding to the first trust degree and a second weight parameter corresponding to the second trust degree, wherein the sum of the first weight parameter and the second weight parameter is a constant value, the first weight parameter is used to adjust the trust degree of the action event in the external interaction information, and the second weight parameter is used to adjust the trust degree of the action event in the internal sensing information; and add a first product of the first trust degree and the first weight parameter and a second product of the second trust degree and the second weight parameter to obtain the evaluation result.

[0128] In an example embodiment, the trust degree evaluation module is further configured to determine a state space set and an action space set corresponding to the environment context information, wherein the action space set comprises a plurality of preset weight parameters, the plurality of preset weight parameters represent an influence degree of the first trust degree in the evaluation result, and the state space set comprises a plurality of state elements; after a current state element is selected from the plurality of state elements each time, the first weight parameter and the second weight parameter are updated in an environment state corresponding to the current state element by the following manner: in a case where the plurality of preset weight parameters are selected for the first time, a first weight parameter with a maximum value is selected from the plurality of preset weight parameters, a selection action occurring when the first weight parameter is selected is determined, and a feedback incentive value set for the selection action is determined, wherein the feedback incentive value is used to represent a trust degree of the selection action, in a case where the feedback incentive value is a positive number, the selection action corresponds to positive feedback, and in a case where the feedback incentive value is a negative number, the selection action corresponds to negative feedback; the feedback incentive value is iteratively calculated using an iterative function to obtain a first incentive value corresponding to the selection action; the action space set is traversed, a weight parameter is selected from the plurality of preset weight parameters multiple times to obtain a plurality of first incentive values; a maximum value in the plurality of first incentive values is taken as a target incentive value, a target weight parameter corresponding to the target incentive value is determined, and the first weight parameter and the second weight parameter are updated using the target weight parameter.

[0129] In an example embodiment, it is to be noted that each group of state elements comprises at least: a first element used to represent a first information entropy of the external interaction information, a second element used to represent a second information entropy of the internal sensing information, a third element used to represent an information relative quantity of the internal sensing information and the external interaction information, and a fourth element used to represent a regularity of a trust value of trust information corresponding to the action event, wherein the first element and the second element are discrete variables, and the trust information is from the internal sensing information and the external interaction information.

[0130] Those skilled in the art can clearly understand that the method according to the above-mentioned embodiments can be realized by means of software and a general hardware platform as required, of course, it can also be realized by hardware, but in many cases, the former is a better implementation. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a readable storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a plurality of instructions for making a terminal device (which can be a mobile phone, computer, server, or network device, etc.) execute the method of each embodiment of the present application.

[0131] In an example embodiment, the computer readable storage medium described above can include, but is not limited to, a U disk, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store computer programs.

[0132] The specific examples in the present embodiment can refer to the examples described in the above embodiments and example embodiments, which will not be repeated here.

[0133] The embodiments of the present application also provide an electronic device, including a memory and a processor, the memory stores a computer program, and the processor is configured to execute the computer program to perform the steps in any of the above method embodiments.

[0134] Optionally, in the present embodiment, the processor can be configured to execute the following steps through the computer program:

[0135] S1, the received second unmanned equipment sends the communication message to analyze and obtain the evaluation request, wherein the evaluation request at least includes the action event to be executed by the first unmanned equipment.

[0136] S2, the historical trust degree of the action event is determined based on the environment context information corresponding to the action event, and the trust degree of the action event is evaluated according to the environment context information, and the evaluation result is obtained, wherein the environment context information at least includes the internal sensing information monitored by the sensing device built in the first unmanned equipment, and the external interaction information generated when the first unmanned equipment and the second unmanned equipment interact, the external interaction information includes the external environment information of the first unmanned equipment and the device message of the second unmanned equipment.

[0137] S3, in the case of determining that the evaluation result is used to indicate that the trust degree of the action event is greater than the preset threshold of the trust degree, the action event is responded.

[0138] In an example embodiment, the electronic device described above can also include a transmission device and an input and output device, wherein the transmission device is connected with the processor, and the input and output device is connected with the processor.

[0139] Optionally, in the present embodiment, the electronic device described above can also be configured to execute the above steps S1 to S3 through the computer program.

[0140] The specific examples in the present embodiment can refer to the examples described in the above embodiments and exemplary embodiments, which will not be repeated here.

[0141] Obviously, those skilled in the art should understand that the modules or steps of the present application described above can be realized by general computing devices, which can be centralized on a single computing device or distributed on a network composed of multiple computing devices, which can be realized by program codes executable by the computing devices, so that they can be stored in storage devices and executed by the computing devices, and in some cases, the steps shown or described can be executed in different order, or they can be respectively manufactured into individual integrated circuit modules, or multiple modules or steps among them can be manufactured into a single integrated circuit module to realize. Thus, the present application is not limited to any specific combination of hardware and software.

[0142] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Those skilled in the art can make various modifications and changes to the present application. Any modification, equivalent replacement, improvement, etc. within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A trust assessment method of an unmanned system, characterized by, Applications to first-class unmanned equipment include: The received communication message sent by the second unmanned device is parsed to obtain an evaluation request, wherein the evaluation request includes at least the action event to be performed by the first unmanned device; Based on the historical trust level of the action event, the environmental context information corresponding to the action event is determined, and the trust level of the action event is evaluated according to the environmental context information to obtain the evaluation result. The environmental context information includes at least the internal sensing information monitored by the built-in sensing device of the first unmanned device, and the external interaction information generated when the first unmanned device and the second unmanned device interact. The external interaction information includes the external environment information of the first unmanned device and the device messages of the second unmanned device. If the evaluation result indicates that the trust level of the action event is greater than a preset trust threshold, then the action event is responded to. Based on the historical trust level of the action event, the environmental context information corresponding to the action event is determined, including: The historical data of the action event in the historical scenario is obtained from the database, wherein the historical data includes at least the historical trust level of the action event, the historical trust level corresponding to the historical trust level, and the time factor corresponding to the historical trust level; Based on the historical event trust level and the time factor, generate the environmental context information corresponding to the action event; Based on the historical event trust level and the time factor, the environmental context information corresponding to the action event is generated, including: Determine the event scenario corresponding to the scene variables of the action event and the historical scene; Determine a first dataset related to the event scenario and corresponding to the external interaction information, and determine a second dataset related to the event scenario and corresponding to the internal sensing information; The first dataset includes a first tuple of the first factor and the second factor of each element corresponding to the external interaction information, and the second dataset includes a second tuple of the third factor and the fourth factor of each element corresponding to the internal sensing information. The first factor and the third factor represent the trust level of the historical event, and the second factor and the fourth factor represent the time factor. The environmental context information is determined based on a first set generated from the first tuple in the first dataset and a second set generated from the second tuple in the second dataset. 2.The trust assessment method of the unmanned system according to claim 1, characterized in that, After retrieving the environmental context information corresponding to the action event from the database, the method further includes: The information format of the environmental context information is converted into a tuple format according to the information conversion rules to obtain the converted environmental context information. The tuple format includes at least the following format parameters: information source parameter, event type parameter, event occurrence location, event trust level, and time factor. The step of evaluating the trust level of the action event based on the environmental context information to obtain the evaluation result includes: According to the converted environmental context information, a trust degree of the action event is evaluated by an evaluation function, and an evaluation result is obtained. 3.The trust assessment method of the unmanned system according to claim 1, wherein, According to the environmental context information, a trust degree of the action event is evaluated by an evaluation function, and an evaluation result is obtained, including: A first information entropy of the first data set is calculated, and a second information entropy of the second data set is calculated; A first trust degree corresponding to the external interaction information is determined based on a first data mean value of the first data set and the first information entropy, and a second trust degree corresponding to the internal sensing information is determined based on a second data mean value of the second data set and the second information entropy, wherein the first data mean value represents a mean value obtained by averaging the numerical values of all elements contained in the first data set, and the second data mean value represents a mean value obtained by averaging the numerical values of all elements contained in the second data set; The first trust degree and the second trust degree are determined as the evaluation result. 4.The trust assessment method of the unmanned system according to claim 3, wherein, The first trust degree and the second trust degree are determined as the evaluation result, including: A first weight parameter corresponding to the first trust degree and a second weight parameter corresponding to the second trust degree are determined, wherein the sum of the first weight parameter and the second weight parameter is a constant value, the first weight parameter is used to adjust the trust degree of the action event in the external interaction information, and the second weight parameter is used to adjust the trust degree of the action event in the internal sensing information; A first product of the first trust degree and the first weight parameter and a second product of the second trust degree and the second weight parameter are added to obtain the evaluation result. 5.The trust assessment method of the unmanned system according to claim 4, characterized in that, The method further includes: A state space set and an action space set corresponding to the environmental context information are determined, wherein the action space set includes a plurality of preset weight parameters, the plurality of preset weight parameters represent the influence degree of the first trust degree in the evaluation result, and the state space set contains a plurality of groups of state elements; After selecting a current state element from the plurality of groups of state elements each time, the first weight parameter and the second weight parameter are updated in an environmental state corresponding to the current state element by the following method: In the case of selecting the plurality of preset weight parameters for the first time, a first weight parameter with the maximum value is selected from the plurality of preset weight parameters, a selection action occurring when the first weight parameter is selected is determined, and a feedback incentive value set for the selection action is determined, wherein the feedback incentive value is used to represent the credibility of the selection action, in the case that the feedback incentive value is a positive number, it represents that the selection action corresponds to positive feedback, and in the case that the feedback incentive value is a negative number, it represents that the selection action corresponds to negative feedback; An iterative function is used to iteratively calculate the feedback incentive value, and a first incentive value corresponding to the selection action is obtained; The action space set is traversed, and weight parameters are selected from the plurality of preset weight parameters multiple times to obtain a plurality of first incentive values; The action space set is traversed, and weight parameters are selected from the plurality of preset weight parameters multiple times to obtain a plurality of first incentive values; A maximum value in the plurality of first excitation values is taken as a target excitation value, and a target weight parameter corresponding to the target excitation value is determined, and the first weight parameter and the second weight parameter are updated using the target weight parameter. 6.The trust assessment method of the unmanned system according to claim 5, wherein, Each group of state elements at least includes: a first element for representing a first information entropy of the external interaction information, a second element for representing a second information entropy of the internal sensing information, a third element for representing an information relative quantity of the internal sensing information and the external interaction information, and a fourth element for representing a regularity of a trust value of trust information corresponding to the action event, wherein the first element and the second element are discrete variables, and the trust information is from the internal sensing information and the external interaction information.

7. A computer readable storage medium, characterized in that, The storage medium has a computer program stored therein, wherein the computer program is configured to execute the method in any one of claims 1 to 6 when running.

8. An electronic device, comprising: The computer program product comprises a memory and a processor, the memory has a computer program stored therein, and the processor is configured to execute the method in any one of claims 1 to 6 by using the computer program.

9. A computer program product comprising computer instructions, characterized in that, The computer program product comprises a memory and a processor, the memory has a computer program stored therein, and the processor is configured to execute the method in any one of claims 1 to 6 by using the computer program. The computer program product comprises a memory and a processor, the memory has a computer program stored therein, and the processor is configured to execute the method in any one of claims 1 to 6 by using the computer program.

Images